Abelian variety
A higher-dimensional relative of an elliptic curve, with a compatible addition rule.
Read the chapter →A visual journey through every subject in the collection. Examples first. Terms explained.
Choose a chapter to jump to its explanation. Use the video controls for fullscreen playback.
The repository README describes 722 manuscripts in 372 families. Families group main theorems, alternative arguments, and consequences. The package also contains manuscript source files, Lean proof artifacts, checking configurations, and ten abridged reasoning summaries. OpenAI reports about 4,000 posed problems and about three hours of Pro-equivalent compute per result on average. The counts alone do not measure a success rate. The README identifies procedural exceptions for zeta and CM Hodge work, and human editing of the 11/12 zeta writeup.
The film visits all 17 subjects and discusses 85 selected families. The index below includes every family and every manuscript in the pinned release. The film is selective within each subject. Its examples explain concepts; they are not reconstructions of the proofs.
The catalogue records differing verification stages and review status “unchecked.” We checked source statements, scopes, counts, and artifact listings for this production. We did not rebuild the Lean project or independently verify the mathematical proofs. A “source listed” badge only records an association in the formalization catalogue. It does not certify the whole paper or family.
Snapshot: adc7f1241b42, checked 7 October 2026. Original catalogue PDF · Manuscript map · Formalization catalogue · Checking instructions
THE COMPLETE COLLECTION
Search all 372 families and 722 papers. Choose a subject, revisit an explanation, or open a family to read its exact scope and sources.
Bars count families, not importance. Each subject appears in the film.
Start with a familiar subject or search for a problem. Technical descriptions retain the repository’s language; the film and glossary provide gentler introductions.
372 families · 722 papers
Number theory
Original repository description
Number theory
Original repository description
Number theory
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Number theory
Original repository description
Number theory
Original repository description
Number theory
Original repository description
Number theory
Original repository description
Number theory
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Number theory
Original repository description
Number theory
Original repository description
Number theory
Original repository description
Number theory
Original repository description
Number theory
Original repository description
Number theory
Original repository description
Number theory
Original repository description
Number theory
Original repository description
Number theory
Original repository description
Number theory
Original repository description
Number theory
Original repository description
Number theory
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Number theory
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Number theory
Original repository description
Number theory
Original repository description
Number theory
Original repository description
Number theory
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Number theory
Original repository description
Number theory
Original repository description
Number theory
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Number theory
Original repository description
Number theory
Original repository description
Number theory
Original repository description
Algebraic and complex geometry
Original repository description
Algebraic and complex geometry
Original repository description
Algebraic and complex geometry
Original repository description
Algebraic and complex geometry
Original repository description
Algebraic and complex geometry
Original repository description
Algebraic and complex geometry
Original repository description
Algebraic and complex geometry
Original repository description
Algebraic and complex geometry
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Algebraic and complex geometry
Original repository description
Algebraic and complex geometry
Original repository description
Algebraic and complex geometry
Original repository description
Algebraic and complex geometry
Original repository description
Algebraic and complex geometry
Original repository description
Algebraic and complex geometry
Original repository description
Algebraic and complex geometry
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Algebraic and complex geometry
Original repository description
Algebraic and complex geometry
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Algebraic and complex geometry
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Algebraic and complex geometry
Original repository description
Algebraic and complex geometry
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Algebraic and complex geometry
Original repository description
Algebraic and complex geometry
Original repository description
Algebraic and complex geometry
Original repository description
Algebraic and complex geometry
Original repository description
Algebraic and complex geometry
Original repository description
Algebraic and complex geometry
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Algebraic and complex geometry
Original repository description
Algebraic and complex geometry
Original repository description
Algebraic and complex geometry
Original repository description
Algebraic and complex geometry
Original repository description
Algebraic and complex geometry
Original repository description
Algebraic and complex geometry
Original repository description
Algebraic and complex geometry
Original repository description
Algebraic and complex geometry
Original repository description
Algebraic and complex geometry
Original repository description
Algebraic and complex geometry
Original repository description
Real and complex analysis
Original repository description
Real and complex analysis
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Real and complex analysis
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Real and complex analysis
Original repository description
Real and complex analysis
Original repository description
Real and complex analysis
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Real and complex analysis
Original repository description
Real and complex analysis
Original repository description
Real and complex analysis
Original repository description
Real and complex analysis
Original repository description
Real and complex analysis
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Real and complex analysis
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Real and complex analysis
Original repository description
Real and complex analysis
Original repository description
Real and complex analysis
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Real and complex analysis
Original repository description
Convex and metric geometry
Original repository description
Read the abridged reasoning summary →
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Convex and metric geometry
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Convex and metric geometry
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Convex and metric geometry
Original repository description
Convex and metric geometry
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Convex and metric geometry
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Convex and metric geometry
Original repository description
Convex and metric geometry
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Convex and metric geometry
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Convex and metric geometry
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Convex and metric geometry
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Convex and metric geometry
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Convex and metric geometry
Original repository description
Convex and metric geometry
Original repository description
Convex and metric geometry
Original repository description
Theoretical computer science
Original repository description
Read the abridged reasoning summary →
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Theoretical computer science
Original repository description
Theoretical computer science
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Theoretical computer science
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Theoretical computer science
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Theoretical computer science
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Theoretical computer science
Original repository description
Theoretical computer science
Original repository description
Theoretical computer science
Original repository description
Theoretical computer science
Original repository description
Theoretical computer science
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Theoretical computer science
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Theoretical computer science
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Theoretical computer science
Original repository description
Theoretical computer science
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Theoretical computer science
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Theoretical computer science
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Theoretical computer science
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Theoretical computer science
Original repository description
Theoretical computer science
Original repository description
Theoretical computer science
Original repository description
Theoretical computer science
Original repository description
Theoretical computer science
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Theoretical computer science
Original repository description
Theoretical computer science
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Theoretical computer science
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Theoretical computer science
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Theoretical computer science
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Theoretical computer science
Original repository description
Theoretical computer science
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Theoretical computer science
Original repository description
Theoretical computer science
Original repository description
Theoretical computer science
Original repository description
Theoretical computer science
Original repository description
Theoretical computer science
Original repository description
Theoretical computer science
Original repository description
Theoretical computer science
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Theoretical computer science
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Theoretical computer science
Original repository description
Theoretical computer science
Original repository description
Dynamical systems and ergodic theory
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Dynamical systems and ergodic theory
Original repository description
Dynamical systems and ergodic theory
Original repository description
Dynamical systems and ergodic theory
Original repository description
Dynamical systems and ergodic theory
Original repository description
Dynamical systems and ergodic theory
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Dynamical systems and ergodic theory
Original repository description
Dynamical systems and ergodic theory
Original repository description
Dynamical systems and ergodic theory
Original repository description
Dynamical systems and ergodic theory
Original repository description
Dynamical systems and ergodic theory
Original repository description
Dynamical systems and ergodic theory
Original repository description
Combinatorics
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Combinatorics
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Combinatorics
Original repository description
Combinatorics
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Combinatorics
Original repository description
Combinatorics
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Combinatorics
Original repository description
Combinatorics
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Combinatorics
Original repository description
Combinatorics
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Combinatorics
Original repository description
Combinatorics
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Combinatorics
Original repository description
Combinatorics
Original repository description
Combinatorics
Original repository description
Combinatorics
Original repository description
Combinatorics
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Combinatorics
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Combinatorics
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Combinatorics
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Combinatorics
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Combinatorics
Original repository description
Combinatorics
Original repository description
Combinatorics
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Combinatorics
Original repository description
Combinatorics
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Combinatorics
Original repository description
Combinatorics
Original repository description
Combinatorics
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Combinatorics
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Combinatorics
Original repository description
Combinatorics
Original repository description
Combinatorics
Original repository description
Combinatorics
Original repository description
Combinatorics
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Combinatorics
Original repository description
Combinatorics
Original repository description
Algebra
Original repository description
Algebra
Original repository description
Algebra
Original repository description
Algebra
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Algebra
Original repository description
Read the abridged reasoning summary →
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Algebra
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Algebra
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Algebra
Original repository description
Algebra
Original repository description
Algebra
Original repository description
Algebra
Original repository description
Algebra
Original repository description
Algebra
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Algebra
Original repository description
Algebra
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Algebra
Original repository description
Algebra
Original repository description
Algebra
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Probability and statistical mechanics
Original repository description
Probability and statistical mechanics
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Probability and statistical mechanics
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Probability and statistical mechanics
Original repository description
Probability and statistical mechanics
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Probability and statistical mechanics
Original repository description
Probability and statistical mechanics
Original repository description
Probability and statistical mechanics
Original repository description
Probability and statistical mechanics
Original repository description
Probability and statistical mechanics
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Probability and statistical mechanics
Original repository description
Probability and statistical mechanics
Original repository description
Probability and statistical mechanics
Original repository description
Probability and statistical mechanics
Original repository description
Probability and statistical mechanics
Original repository description
Probability and statistical mechanics
Original repository description
Probability and statistical mechanics
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Probability and statistical mechanics
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Probability and statistical mechanics
Original repository description
Probability and statistical mechanics
Original repository description
Probability and statistical mechanics
Original repository description
Probability and statistical mechanics
Original repository description
Probability and statistical mechanics
Original repository description
Probability and statistical mechanics
Original repository description
Probability and statistical mechanics
Original repository description
Probability and statistical mechanics
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Probability and statistical mechanics
Original repository description
Probability and statistical mechanics
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Probability and statistical mechanics
Original repository description
Mathematical logic
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Mathematical logic
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Mathematical logic
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Mathematical logic
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Mathematical logic
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Mathematical logic
Original repository description
Group theory
Original repository description
Group theory
Original repository description
Group theory
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Group theory
Original repository description
Group theory
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Group theory
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Group theory
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Group theory
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Group theory
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Group theory
Original repository description
Group theory
Original repository description
Group theory
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Group theory
Original repository description
Group theory
Original repository description
Mathematical physics
Original repository description
Mathematical physics
Original repository description
Mathematical physics
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Mathematical physics
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Mathematical physics
Original repository description
Mathematical physics
Original repository description
Mathematical physics
Original repository description
Mathematical physics
Original repository description
Mathematical physics
Original repository description
Mathematical physics
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Mathematical physics
Original repository description
Mathematical physics
Original repository description
Mathematical physics
Original repository description
Mathematical physics
Original repository description
Mathematical physics
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Mathematical physics
Original repository description
Mathematical physics
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Mathematical physics
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Mathematical physics
Original repository description
Mathematical physics
Original repository description
Mathematical physics
Original repository description
Mathematical physics
Original repository description
Mathematical physics
Original repository description
Mathematical physics
Original repository description
Mathematical physics
Original repository description
Operator algebras
Original repository description
Operator algebras
Original repository description
Operator algebras
Original repository description
Operator algebras
Original repository description
Operator algebras
Original repository description
Operator algebras
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Operator algebras
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Operator algebras
Original repository description
Operator algebras
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Operator algebras
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Operator algebras
Original repository description
Operator algebras
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Operator algebras
Original repository description
Operator algebras
Original repository description
Operator algebras
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Operator algebras
Original repository description
Operator algebras
Original repository description
Operator algebras
Original repository description
Operator algebras
Original repository description
Topology
Original repository description
Topology
Original repository description
Topology
Original repository description
Topology
Original repository description
Topology
Original repository description
Topology
Original repository description
Topology
Original repository description
Topology
Original repository description
Topology
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Topology
Original repository description
Topology
Original repository description
Topology
Original repository description
Topology
Original repository description
Topology
Original repository description
Topology
Original repository description
Topology
Original repository description
Topology
Original repository description
Topology
Original repository description
Functional analysis
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Functional analysis
Original repository description
Functional analysis
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Functional analysis
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Functional analysis
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Functional analysis
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Functional analysis
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Functional analysis
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Functional analysis
Original repository description
Functional analysis
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Functional analysis
Original repository description
Differential geometry
Original repository description
Differential geometry
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Differential geometry
Original repository description
Differential geometry
Original repository description
Differential geometry
Original repository description
Differential geometry
Original repository description
Differential geometry
Original repository description
Differential geometry
Original repository description
Differential geometry
Original repository description
Differential geometry
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Differential geometry
Original repository description
Differential geometry
Original repository description
Differential geometry
Original repository description
Differential geometry
Original repository description
Differential geometry
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Differential geometry
Original repository description
Differential geometry
Original repository description
Differential geometry
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Differential geometry
Original repository description
Differential geometry
Original repository description
Differential geometry
Original repository description
Differential geometry
Original repository description
Differential geometry
Original repository description
Differential geometry
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Differential geometry
Original repository description
Differential geometry
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Differential geometry
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Differential geometry
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Differential geometry
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Partial differential equations
Original repository description
Read the abridged reasoning summary →
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Partial differential equations
Original repository description
Partial differential equations
Original repository description
Partial differential equations
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Partial differential equations
Original repository description
Partial differential equations
Original repository description
Partial differential equations
Original repository description
Partial differential equations
Original repository description
Partial differential equations
Original repository description
“Source listed” identifies a paper in formalization.yaml. It is an artifact association, not a verification certificate for this family.
Partial differential equations
Original repository description
Partial differential equations
Original repository description
Partial differential equations
Original repository description
Partial differential equations
Original repository description
Partial differential equations
Original repository description
Partial differential equations
Original repository description
Partial differential equations
Original repository description
No matching families. Try fewer words or reset the filters.
Repository descriptions are reproduced from OpenAI’s catalogue under Apache License 2.0. They state the authors’ results and assumptions. Read the source catalogue · Source license.
Look up a term, then return to the chapter that introduces it. These short explanations are starting points; exact conditions are in the chapter notes.
139 terms
A higher-dimensional relative of an elliptic curve, with a compatible addition rule.
Read the chapter →A formal combination of geometric pieces cut out by polynomial equations.
Read the chapter →A precise sequence of instructions a computer can follow.
Read the chapter →For finitely generated groups, large finite pieces can have boundary-to-size ratio as small as desired.
Read the chapter →Rank 0: nonzero at 1. Rank 1: a simple zero, changing roughly linearly near 1.
Read the chapter →A lower bound on how good the algorithm’s answer is relative to the best answer.
Read the chapter →Entanglement is bounded in proportion to a region’s boundary length, under the stated gap and interaction assumptions.
Read the chapter →Area of the body multiplied by area of its polar.
Read the chapter →A sequence with the same gap between consecutive terms.
Read the chapter →About growth as the problem size becomes sufficiently large.
Read the chapter →A normed vector space containing the limit of every sequence whose terms become arbitrarily close to one another.
Read the chapter →A point where all sections in the chosen system vanish, so they cannot provide coordinates there.
Read the chapter →Family 032 also covers finite-field consequences and products of projective K3 surfaces.
Read the chapter →Its vertices split into two classes, with every edge joining different classes.
Read the chapter →Count elementary operations on bits, the zeros and ones encoding the input.
Read the chapter →A nonvanishing fraction of all particles occupy one quantum state in the large-system limit.
Read the chapter →An ideal collection of directions in which paths can escape indefinitely.
Read the chapter →The representation preserves every distance within a fixed multiplicative factor after rescaling.
Read the chapter →If X × line and Y × line agree, must X and Y agree? The reported example says no.
Read the chapter →All models of the specified size are equivalent through a structure-preserving match.
Read the chapter →ab = 1, ac = 0, and c ≠ 0 rule out ba = 1.
Read the chapter →Adding 1 to itself gives 0.
Read the chapter →A locally straight path that closes up smoothly. Distinct images are different geometric paths.
Read the chapter →One element in an algebraic description of a space’s global structure.
Read the chapter →Many compatible ways of sending the space to itself while respecting addition. Precise condition in the notes.
Read the chapter →A number with real and imaginary coordinates, shown as a point on a plane.
Read the chapter →A map preserving angles locally, though lengths and areas can change.
Read the chapter →A mathematical statement suspected to be true.
Read the chapter →The partial totals approach a finite number.
Read the chapter →Every straight segment joining two points of the filled shape stays inside it.
Read the chapter →A specific construction for which the proposed universal rule fails.
Read the chapter →The transition value for the appearance of an infinite open cluster.
Read the chapter →The smallest possible number of edge crossings among all allowed plane drawings of a graph.
Read the chapter →Degree measures polynomial complexity; multiplicity measures repeated vanishing at a point.
Read the chapter →The integer below the fraction bar. Larger q allows finer choices of fractions.
Read the chapter →The proportion of exceptions tends to zero in the specified ordering.
Read the chapter →Replace random choices by a deterministic procedure while preserving the stated resource bounds.
Read the chapter →Theorems, counterexamples, sharper bounds, constructions, and links between theories.
Read the chapter →Each spin interacts with a finite average number of other spins.
Read the chapter →The property that ab = 1 always forces ba = 1.
Read the chapter →Particle density as a function of position and momentum.
Read the chapter →Its positive partial totals grow without bound.
Read the chapter →Multiply corresponding coordinates, then add: (a,b)·(x,y) = ax + by.
Read the chapter →The minimum number of allowed one-letter edits converting one word into another.
Read the chapter →A sum of distinct positive fractions of the form 1/n.
Read the chapter →A curve such as y² = x³ − x, completed by a point at infinity, with a rule for adding points.
Read the chapter →Compare arrangements at the same density using the same pair-interaction rule.
Read the chapter →A positive allowance, however small, fixed before matrix size grows.
Read the chapter →A solution exists, and the same initial state determines only one solution.
Read the chapter →The power of n controlling the growth of the arithmetic count.
Read the chapter →The precise bound allows n raised to 2.25 plus any fixed positive extra amount.
Read the chapter →A system whose interactions favor magnetic alignment.
Read the chapter →A ratio comparing particle number with the number of available magnetic states.
Read the chapter →An arithmetic system with finitely many values in which every nonzero value has a multiplicative inverse.
Read the chapter →A finite list of generators and relations describing a group, which can still be infinite.
Read the chapter →A point x satisfying T(x) = x. A nonexpansive T never increases the distance between two points.
Read the chapter →An approximation formed by keeping finitely many frequency components.
Read the chapter →A measure of how the number of small pieces needed to cover a set grows as their size shrinks.
Read the chapter →A quantity balancing energy and how many arrangements are available at a given temperature.
Read the chapter →Reduced words stay distinct; a letter cancels its adjacent inverse.
Read the chapter →A prime in the arithmetic of numbers a + bi with integer a and b.
Read the chapter →A basic group element from which other elements are made using products and inverses.
Read the chapter →An element from which the whole algebra is obtained using its permitted operations and limits.
Read the chapter →Features of the whole space, such as the way holes occur.
Read the chapter →Vertices are dots. Edges are the connections between them.
Read the chapter →A system of reversible operations. They can be combined, and each can be undone.
Read the chapter →Add weighted group elements. Multiply by distributing, then combining the group operations.
Read the chapter →A spatial temperature pattern whose shape stays fixed while its amplitude decays during diffusion.
Read the chapter →A homeomorphism is a reversible continuous map. Homotopy equivalence permits inverse maps only up to continuous deformation.
Read the chapter →A subspace preserved by every operator that commutes with the chosen operator.
Read the chapter →Other researchers examine the statements, arguments, and supporting evidence.
Read the chapter →A starting spatial geometry, its rate of change, and matter or field information satisfying the required constraints.
Read the chapter →A weighted algebraic count of an intersection, which includes more information than simply counting visible points.
Read the chapter →A measure of how often rational approximations can achieve exceptionally small errors.
Read the chapter →A smooth realization preserving lengths of tangent vectors, while allowing self-intersections.
Read the chapter →Related by a one-to-one correspondence preserving the mathematical structure.
Read the chapter →A lower bound on boundary size for a region of given volume.
Read the chapter →A set containing a unit line segment in every direction.
Read the chapter →Remove a tubular neighborhood of a knot, then reglue it with a chosen boundary identification.
Read the chapter →A function packaging arithmetic information, here from counts of solutions modulo primes.
Read the chapter →Add the absolute differences of the coordinates: |x₁ − y₁| + |x₂ − y₂| + …
Read the chapter →Software that checks a precisely written mathematical argument.
Read the chapter →An isolated periodic orbit: a repeating trajectory separated from other repeating trajectories.
Read the chapter →+1 for an even number of prime factors; −1 for an odd number, counting repeats.
Read the chapter →The net magnetic effect per unit size, after adding the individual magnetic moments.
Read the chapter →A space that looks like ordinary Euclidean space in a sufficiently small neighborhood of every point.
Read the chapter →A selection of edges with no shared endpoints.
Read the chapter →A rectangular table of numbers.
Read the chapter →Long time separation makes the relevant correlations approach products of probabilities.
Read the chapter →A measure of motion; at low speeds, mass multiplied by velocity.
Read the chapter →A partner that multiplies with an element to give 1, the multiplicative identity.
Read the chapter →One kind of charged particle. The full starting-state conditions are in the guide.
Read the chapter →The value depends only on one linear coordinate, even when the surrounding space has many dimensions.
Read the chapter →A linear rule acting on vectors, possibly in a space with infinitely many coordinates.
Read the chapter →The largest factor by which an operator can stretch a unit vector.
Read the chapter →The largest crossing-edge count achievable by any split.
Read the chapter →If a point is in the body, its coordinatewise negative is too.
Read the chapter →Does efficient checking always imply efficient solving? P ≠ NP says no.
Read the chapter →One family can contain several related manuscripts.
Read the chapter →A lower bound on mass in terms of the minimum area enclosing a trapped region, under precise geometric conditions.
Read the chapter →After enough time, all concentrations stay bounded away from both zero and infinity, within the stated class.
Read the chapter →Matter containing freely moving charged particles.
Read the chapter →The points whose dot product with every point of the original body is at most 1.
Read the chapter →A finite sum of terms using nonnegative whole-number powers of the variables.
Read the chapter →Its powers visit all nonzero remainders modulo the prime.
Read the chapter →An argument deriving a statement from stated assumptions.
Read the chapter →Curves obtained by a controlled change of coefficients; one form is d y² = x³ + ax + b.
Read the chapter →The number of independent starting points needed for its infinite part, using repeated addition.
Read the chapter →Rational: an exact fraction of integers. Irrational: no such exact fraction exists.
Read the chapter →A solution whose coordinates are fractions of integers.
Read the chapter →One divided by the number: the reciprocal of 7 is 1/7.
Read the chapter →If the starting axioms are consistent, a model with the claimed combination of properties can be built.
Read the chapter →Count the stages needed to describe an algebraic object through successive equations and relations.
Read the chapter →The limiting transition between constraints that can all be satisfied and constraints that cannot.
Read the chapter →A large-system limit obtained while rescaling distances, time, or other quantities.
Read the chapter →Change coordinates by an invertible transformation: A becomes S A S⁻¹.
Read the chapter →A failure of the smooth solution, for example through uncontrolled growth of derivatives.
Read the chapter →A quantum property that can make particles behave like tiny magnets.
Read the chapter →A disordered spin system with interactions that can favor conflicting alignments.
Read the chapter →No prime square divides the number. For example, 10 qualifies; 12 does not.
Read the chapter →A question about whether physically admissible spacetime evolution can continue past the region determined by initial data.
Read the chapter →Each bar counts result families assigned to that subject by the catalogue.
Read the chapter →Papers, source files, formal proof artifacts, checking instructions, and selected reasoning summaries.
Read the chapter →The region with horizontal coordinate strictly between 0 and 1.
Read the chapter →T^(3/2) = T × √T. The theorem also specifies the multiplier.
Read the chapter →A level of computational power, defined by mutual ability to solve problems using each other as an oracle.
Read the chapter →Multiply signs a fixed distance apart, then average those products.
Read the chapter →The separation above the lowest energy stays bounded away from zero as the system size grows.
Read the chapter →Labeling puzzles with pairwise rules: one endpoint’s label uniquely determines the required other label.
Read the chapter →Membership is represented by exactly one auxiliary solution, and nonmembership by none.
Read the chapter →Points at distance exactly 1 must receive different colors.
Read the chapter →A system can simulate any computation when supplied with an appropriate encoded program and input.
Read the chapter →One algorithm that always finishes and gives the correct yes-or-no answer.
Read the chapter →Different small-scale models share the same limiting behavior under the stated conditions.
Read the chapter →A collection of vector transformations that also contains the relevant combinations and limits.
Read the chapter →A solution satisfying the equations in an integrated sense, allowing less smoothness than a classical solution.
Read the chapter →A nonzero element a for which ab = 0 for some nonzero b.
Read the chapter →An input for which the function’s output equals zero.
Read the chapter →A region containing no inputs where the function vanishes.
Read the chapter →The complete narration, plain-language definitions, scope notes and primary sources.
This release reaches far beyond prime numbers and a few famous conjectures. It asks how shapes can be bent, whether randomness helps computers, how gases evolve, and what quantum systems can do. This expanded tour visits every one of the catalogue's seventeen subjects. We will explain representative results through examples and define the unfamiliar words as they appear. The companion page lets you search every result family and follow every paper.
The package contains seven hundred twenty-two manuscripts, grouped into three hundred seventy-two result families. A family can include a main theorem, another proof, and papers developing its consequences. A manuscript is one paper; a family is a group of related papers. The numbers are not interchangeable.
There are also manuscript source files, a Lean proof library, checking configurations, and abridged reasoning summaries for ten selected families. Those summaries describe parts of the reasoning process; they are not substitutes for the mathematical proofs. OpenAI says the model was posed about four thousand problems. The published counts alone do not give a success rate.
A conjecture is a mathematical statement people suspect is true.
A proof is an argument establishing that it follows from stated assumptions. The release includes written proofs and, for many results, versions intended to be checked by software called Lean.
Lean checks whether the formal proof follows its rules. People still need to check that the formal statement expresses the intended mathematics.
Independent review means other researchers examine that evidence. The results are at different stages of checking; the catalogue marks review status unchecked. So throughout this film, reported result means what the repository claims, not a proof that we have independently verified.
The formalization catalogue reports scope “Partial progress” and review status “unchecked.” Presence in this catalogue means supporting artifacts are listed. It does not mean this production rebuilt the proof library or completed an independent mathematical review.
Here is the full map. The largest family count is in theoretical computer science, but substantial groups of results also concern algebraic geometry, combinatorics, probability, and differential geometry. A family count measures how the catalogue is organized, not how difficult or important a theorem is.
Some results establish a proposed rule. Others give counterexamples: constructions showing that a proposed rule can fail. Still others improve a numerical bound or connect two descriptions of the same mathematical object. We will keep those different kinds of accomplishment visible.
Start with a function: a rule that takes an input and returns an output. A zero is an input where the output is zero. The zeta function is a special function tied to prime numbers, the integers greater than one divisible only by themselves and one.
Its inputs can be complex numbers. Picture those as points on a plane: a real coordinate horizontally, and an imaginary coordinate vertically.
The zeros relevant here lie in the strip between horizontal coordinates zero and one. The Riemann hypothesis says they all sit on the one-half line.
The release reports that none lie to the right of seven eighths. That forbidden region continues upward and downward without end. This is progress toward the famous conjecture, not the full conjecture. The paper extends the claim to related functions used to study primes.
The zero-free region is Re(s) > 7/8, not the full Riemann hypothesis. The principal pole at s = 1 is not a zero. The marked zeta zeros use the first seven positive ordinates, rounded to six decimals. Lean declaration: OAI.riemannZeta_ne_zero_of_seven_eighths_lt_re. The corresponding comparator permits propext, Quot.sound and Classical.choice; this production did not run it.
The following descriptions are reproduced from OpenAI’s catalogue. They report the authors’ claims and retain the technical scope. The film does not independently certify their proofs.
Proves that every Dirichlet -function, including , is zero-free in , resolving the quasi-Riemann hypothesis. The same half-plane is zero-free for every finite-order Hecke -function over . A companion gives a different proof of the zero-free half-plane .
A rational number is a fraction of integers, such as twenty-two sevenths. Pi is irrational: no such fraction equals it exactly. But fractions can come close. Three hundred fifty-five over one hundred thirteen is much better than twenty-two sevenths.
The denominator, the number underneath, is the price paid for that accuracy. Write it as q. An error like one over q squared improves as q grows; one over q cubed would improve faster.
The paper reports that the long-run dividing exponent for pi is two. Choose any fixed exponent larger than two, even two point zero zero one. Only finitely many fractions beat that accuracy target. This statement is called an irrationality exponent of two.
A related infinite sum, the Flint Hills series, then converges: its partial totals approach a finite number. A separate paper reports that another particular infinite sum, Catalan's constant, cannot equal a fraction either.
μ(π) = 2 means that for every ε > 0, all sufficiently large q satisfy |π − p/q| ≥ q^(−2−ε) for every integer p. The two displayed absolute approximation errors are computed from 22/7 and 355/113. The conclusion concerns sufficiently large denominators; it does not give a uniform c/q² lower bound. The Catalan irrationality claim is a separate manuscript.
The following descriptions are reproduced from OpenAI’s catalogue. They report the authors’ claims and retain the technical scope. The film does not independently certify their proofs.
Proves that the irrationality exponent of is exactly : for every and all sufficiently large denominators , every rational satisfies . This also proves convergence of the Flint–Hills series , with angles in radians.
Proves that Catalan's constant is irrational.
Now ask for exact solutions instead of approximations. In x squared plus y squared equals one, x equals three fifths and y equals four fifths work. They are rational solutions because both coordinates are fractions. In x squared equals two, neither solution is rational.
An algorithm is a precise, step-by-step recipe a computer can follow. Could one recipe decide whether every polynomial equation has a rational solution?
Here polynomial means a sum of terms made from whole-number powers of the variables, and the coefficients are integers.
The reported answer is no, when any number of variables is allowed. This is Hilbert's tenth problem over the rationals. It does not say every individual equation is impossible to solve. It says no single procedure can always finish and answer correctly for every input.
The polynomial has integer coefficients, the requested zeros are rational, and the number of variables is part of the input. The result concerns a universal algorithm that always halts, not the solvability of every particular equation.
The following descriptions are reproduced from OpenAI’s catalogue. They report the authors’ claims and retain the technical scope. The film does not independently certify their proofs.
Proves that no algorithm decides whether an integer-coefficient polynomial in an arbitrary number of variables has a rational zero, resolving Hilbert's tenth problem over negatively.
An elliptic curve is a special kind of curve with no sharp corners or crossing singularities, described by an equation like y squared equals x cubed minus x, together with one point at infinity. Despite the name, it is not an ellipse. We want its rational points: solutions whose coordinates are fractions.
There is a rule for adding these points. The curve's rank counts independent starting points needed to generate its infinite part by repeated addition. Think of counting independent directions, while keeping track of a separate finite part.
Another description starts by counting solutions in arithmetic modulo primes. Modulo means keeping only the remainder after division. Those counts are packaged into a function called an L-function. The Birch and Swinnerton-Dyer conjecture links these two descriptions: rational points and the behavior of that function near input one.
The release reports its full numerical formula for a precisely specified class of curves. The extra arithmetic condition is in the notes; this does not settle every curve.
Related work studies families of modified equations called quadratic twists.
It reports a limiting half-and-half split: in half, the L-function is nonzero at one; in half, it has a simple zero, meaning that near one its value changes roughly in proportion to the distance from one. These are called analytic ranks zero and one.
The BSD hypothesis concerns the full q-power Selmer corank, not only the Mordell–Weil rank. Goldfeld densities are over signed squarefree twist parameters ordered by absolute value. The displayed curve y² = x³ − x is an illustration; the three marked rational points are (−1,0), (0,0), and (1,0).
The following descriptions are reproduced from OpenAI’s catalogue. They report the authors’ claims and retain the technical scope. The film does not independently certify their proofs.
Proves the full Birch–Swinnerton-Dyer leading-term formula for every elliptic curve over whose full -power Selmer group has corank zero or one for some prime , including finiteness of the Tate–Shafarevich group. With result 006, this gives full BSD for a density-one set of quadratic twists of every elliptic curve over .
Proves Goldfeld's conjecture for quadratic twists of every elliptic curve over : analytic ranks zero and one each have density , and the mean analytic rank tends to . Both statements order signed squarefree twist parameters by absolute value.
The number-theory package goes further than the preceding examples. The BSD and Goldfeld families are reported to combine into the full BSD formula for almost every quadratic twist of any fixed elliptic curve. Here almost every means that the fraction of exceptions tends to zero as larger twist parameters are included.
Another result concerns squarefree numbers, which have no repeated prime factor. Ten is squarefree; twelve is not, because four divides it. For irreducible fourth-degree integer polynomials without an unavoidable prime-square divisor, the package reports the predicted positive density of squarefree values.
An Egyptian fraction writes a number as a sum of distinct fractions with numerator one. For example, two thirds is one half plus one sixth. The release gives an optimal-order bound on how many terms are needed: the worst case grows like the logarithm of the logarithm of the denominator.
Gaussian primes are primes in arithmetic with integer real and imaginary coordinates. The Gaussian moat result says that no fixed maximum step length lets you walk forever through distinct Gaussian primes.
Finally, a primitive root generates every nonzero remainder modulo a prime by taking powers. Powers of three do this modulo seven. The Artin result reports infinitely many suitable primes for every integer base that is neither minus one nor a square. It proves this infinitude claim, not the full predicted density.
The following descriptions are reproduced from OpenAI’s catalogue. They report the authors’ claims and retain the technical scope. The film does not independently certify their proofs.
Proves the full Birch–Swinnerton-Dyer leading-term formula for every elliptic curve over whose full -power Selmer group has corank zero or one for some prime , including finiteness of the Tate–Shafarevich group. With result 006, this gives full BSD for a density-one set of quadratic twists of every elliptic curve over .
Proves Goldfeld's conjecture for quadratic twists of every elliptic curve over : analytic ranks zero and one each have density , and the mean analytic rank tends to . Both statements order signed squarefree twist parameters by absolute value.
Proves the squarefree-values conjecture for irreducible integer quartics with no fixed prime-square divisor: squarefree values on positive integers have the predicted positive Euler-product density. More generally, establishes the -power-free density for irreducible integer polynomials of degrees four through eight under the necessary local condition; together with Browning's higher-degree theorem, this covers every .
Every rational with is a sum of distinct positive unit fractions. The worst-case minimum number of terms has the same order, resolving Erdős’s conjecture on short Egyptian fractions.
Proves the Gaussian moat conjecture: no infinite walk through distinct Gaussian primes can have uniformly bounded steps. More strongly, for every distance bound , the graph joining Gaussian primes at distance at most has uniformly bounded finite component sizes, depending only on , including primes on the coordinate axes.
Proves the infinitude assertion in Artin's primitive root conjecture for every integer that is neither nor a square. For each such base, at least primes in every sufficiently large interval have primitive root , with .
Imagine studying a shape in two ways. You can describe its global structure, including holes. Or you can describe smaller pieces cut out by equations.
Cohomology is an algebraic way of recording global features; a cohomology class is one item in that record.
An algebraic cycle is a combination of geometric pieces defined by polynomial equations. The Hodge conjecture asks whether certain cohomology classes can be built from these algebraic cycles. The word certain matters: it does not make that claim for every class.
The release reports the conjecture for a special class of spaces called abelian varieties with complex multiplication.
An abelian variety is a higher-dimensional relative of an elliptic curve, with a compatible rule for adding points.
Complex multiplication means it has a particularly rich system of compatible multiplication maps: rules sending the space back to itself while respecting addition. The picture of a doughnut is a visual aid, not the general space. The result covers all dimensions within this special class; it does not settle the Hodge conjecture for every algebraic space.
The principal result concerns rational Hodge classes on complex CM abelian varieties. It is not the general Hodge conjecture. The torus and loops are conceptual illustrations and should not be read as a representation of every algebraic cycle.
The following descriptions are reproduced from OpenAI’s catalogue. They report the authors’ claims and retain the technical scope. The film does not independently certify their proofs.
Proves the rational Hodge conjecture for every complex CM abelian variety, in every dimension and codimension. Through Milne's theorems, this also gives the Tate conjecture for all abelian varieties over finite fields and the Hodge standard conjecture for abelian varieties in every characteristic. Companion results prove rational Hodge for arbitrary products of projective complex K3 surfaces and algebraicity of the Kuga–Satake correspondence for every such surface.
The Hodge family also reports consequences for abelian varieties over finite fields, where arithmetic has only finitely many values. These include the Tate conjecture in that setting. Companion papers extend the rational Hodge claim to products of projective K3 surfaces, special spaces with two complex dimensions, or four real dimensions. These extensions are part of the package, not just the CM case shown earlier.
Other families ask when a shape admits useful coordinate maps built from algebraic data. Think of sections as generalized functions used to build coordinates. A base point is a point where all the available sections vanish together, preventing that coordinate description. Fujita gives a sharp positivity bound that removes base points. Log abundance turns a weaker positivity condition on specified canonical data into such a usable map after taking a suitable multiple. Its proof uses the companion work on how complexity behaves in families of shapes. The exact hypotheses are in the catalogue.
Nagata concerns the cost of forcing a polynomial curve through many points. The degree measures the polynomial complexity; multiplicity measures repeated vanishing at a point. The claimed inequality puts a sharp restriction on their tradeoff for at least ten very general complex points.
The package also contains failures of expected rules. Cancellation would let us remove the same extra coordinate from two equivalent products. A reported complex four-dimensional shape becomes ordinary five-dimensional affine space after adding a line, even though the original shape is not ordinary four-dimensional affine space. Adding a coordinate can hide a difference.
The following descriptions are reproduced from OpenAI’s catalogue. They report the authors’ claims and retain the technical scope. The film does not independently certify their proofs.
Proves the rational Hodge conjecture for every complex CM abelian variety, in every dimension and codimension. Through Milne's theorems, this also gives the Tate conjecture for all abelian varieties over finite fields and the Hodge standard conjecture for abelian varieties in every characteristic. Companion results prove rational Hodge for arbitrary products of projective complex K3 surfaces and algebraicity of the Kuga–Satake correspondence for every such surface.
Proves Campana's orbifold Iitaka subadditivity conjecture for smooth Fujiki-class- manifolds with rational simple-normal-crossing boundaries. For projective fibrations of smooth complex quasi-projective varieties with connected fibers, general fiber , and , proves Popa's inequality , where variation measures the whole geometric generic fiber.
Using logarithmic Iitaka subadditivity, proves log abundance in every dimension for normal compact Kähler log canonical pairs with effective rational boundary: an analytically nef -Cartier adjoint is semiample. It also proves projective log abundance over every algebraically closed field of characteristic zero and the effective Iitaka fibration conjecture for smooth projective varieties of nonnegative Kodaira dimension in that setting. A further result resolves the finite-rational-coefficient index conjecture for connected projective semi-log-canonical log Calabi–Yau pairs over such fields in each fixed dimension and for each fixed finite set of rational boundary coefficients, with a uniform index independent of the number of components.
Proves Fujita's freeness conjecture at its sharp bound in every dimension: for a smooth projective complex variety of dimension and an ample line bundle , the adjoint is globally generated for every integer .
Proves Nagata's strict inequality for every nonzero effective plane curve of degree through very general complex points, with arbitrary multiplicities . It also proves maximal multipoint Seshadri constants for every smooth polarized projective variety of dimension and all sufficiently large : at very general points over , and at the geometric generic tuple over any algebraically closed field of positive characteristic.
Constructs an integral complex affine fourfold with , disproving affine-space cancellation over in dimension four. It also disproves the Dolgachev–Weisfeiler affine-fibration conjecture: smooth surjections and have every residue-field fiber isomorphic to affine three-space but are not Zariski-locally trivial.
Analysis studies functions and limiting processes. Koebe asks whether complicated planar regions can be redrawn while preserving local angles so that the missing pieces become round disks or points. Such an angle-preserving redraw is called conformal. The package reports this existence theorem for every domain on the sphere used in complex analysis.
A fractal dimension measures how finely a set fills space and need not be a whole number. Falconer asks how many different distances a set determines. The reported theorem says that a compact set of dimension greater than half the surrounding dimension has distances occupying positive length on the number line.
Kakeya asks how small a set can be while containing a unit line segment pointing in every direction. The release reports the full four-dimensional dimension conclusion, along with a stronger quantitative tube estimate in three dimensions. Full dimension does not mean positive volume.
Fourier analysis rebuilds a signal from sine and cosine waves. The Fourier result gives pointwise reconstruction except on a set of measure zero under the size condition called L log L. That condition controls large spikes through a finite integral of the function’s size times the logarithm of one plus its size.
The following descriptions are reproduced from OpenAI’s catalogue. They report the authors’ claims and retain the technical scope. The film does not independently certify their proofs.
Resolves the existence part of Koebe's circle-domain conjecture: every domain in the Riemann sphere is conformally equivalent to a domain whose complementary components are round disks or points. It also proves that circle domains with conformally removable boundary are rigid, meaning every conformal equivalence to another circle domain is Möbius, establishing this direction of the He–Schramm conjecture.
Resolves the Falconer distance conjecture in every dimension : every compact set with Hausdorff dimension greater than determines a set of Euclidean distances of positive Lebesgue measure.
Resolves the Kakeya maximal conjecture in three dimensions and the Hausdorff-dimension conjecture in four. In three dimensions, the radius- tube maximal operator maps to with norm for every . In four dimensions, every set containing a unit segment in every direction has Hausdorff dimension four.
Proves that the ordinary symmetric Fourier partial sums of every complex-valued function in converge almost everywhere along the full sequence. This resolves the classical sufficiency conjecture at the scale.
A filled shape is convex if the straight segment between any two of its points stays inside it. A disk and a square are convex; a crescent is not.
Now pair a convex shape with the origin strictly inside it with its polar. The polar consists of the points whose dot product with every point in the original shape is at most one.
A dot product multiplies matching coordinates and adds the results. This rule makes the two shapes respond in opposite ways.
Start with a unit disk; its polar is another unit disk. Each area is pi, so their product is pi squared. As these true polar pairs become a square and a diamond, their area product falls to eight.
The symmetric Mahler claim says this sort of minimum has a sharp formula in every dimension. Symmetric here means unchanged when every coordinate is negated. The equality shapes are called Hanner polytopes, a family that includes cubes and their polar shapes. A companion reports the corresponding result without that symmetry, where the equality shapes are triangles and their higher-dimensional relatives.
Symmetric bound: |K||K°| ≥ 4^n/n!, with equality at invertible linear images of Hanner polytopes. General bound at the Santaló point s(K): |K||(K−s(K))°| ≥ (n+1)^(n+1)/(n!)², with equality at simplices. The animation uses unit ℓᵖ balls and their ℓᑫ polar duals, 1/p + 1/q = 1. The area readout uses the exact gamma-function area formula and ends at the exact square/diamond product 8. The catalogue entries cited cover the symmetric result and its equality classification; the general manuscript is not listed there.
The following descriptions are reproduced from OpenAI’s catalogue. They report the authors’ claims and retain the technical scope. The film does not independently certify their proofs.
Resolves the symmetric and nonsymmetric geometric Mahler conjectures in every dimension, with Hanner polytopes and simplices as the respective volume-product minimizers and all equality cases classified. The corresponding sharp functional Mahler inequalities also hold. For , every symmetric polar product in dimension has Gromov width .
Geometry also asks whether we can change a representation without badly changing distances. In an L-one space, distance is the sum of coordinate differences in absolute value, like traveling along a rectangular street grid. Distortion measures the worst multiplicative mismatch after rescaling.
For shortest-path distances on planar graphs, the package reports an L-one representation with a distortion bound independent of graph size. Planar means the graph can be drawn without edges crossing. A related bound applies to graphs with limited branching complexity called bounded treewidth.
A second result concerns the triangular lattice, the familiar arrangement in which each point has six nearest neighbors. At fixed density, it is reported to minimize energy for a broad specified class of repulsive pair interactions, as well as the stated Coulomb and Riesz energy models. The allowed interaction class matters; this is not a claim about every possible force.
Edit distance counts the cheapest sequence of insertions, deletions, and substitutions converting one word into another. The package determines the sharp growth rate of distortion needed to represent these distances in L-one. So distances that are easy to define need not be easy to reproduce faithfully as coordinate distances.
The following descriptions are reproduced from OpenAI’s catalogue. They report the authors’ claims and retain the technical scope. The film does not independently certify their proofs.
Resolves the planar and bounded-treewidth cases of the Gupta–Newman–Rabinovich–Sinclair conjecture. Shortest-path metrics of finite connected graphs with arbitrary positive edge lengths embed into real with universal distortion for planar graphs, and distortion depending only on treewidth for bounded-treewidth graphs. The corresponding multicommodity flow–cut gaps are uniformly bounded.
Proves that the triangular lattice minimizes the lower limit of energy per particle for every nonnegative completely monotone potential of squared distance among locally finite planar configurations of centered-disk density one. It also minimizes unit-background renormalized Riesz energies for and Coulomb energy, resolving Sandier–Serfaty and the two-dimensional Brauchart–Hardin–Saff conjecture on the linear term of optimal spherical logarithmic energy.
Determines the least distortion of embedding edit distance on words of length at most into real : it is . Insertions, deletions and substitutions have unit cost. The constants are uniform over all finite alphabets with at least two symbols, even when the alphabet grows with ; binary words already force the lower bound.
A graph is a collection of dots connected by lines. The dots are vertices; the lines are edges. In Max-Cut, divide the dots into two groups and count the lines that cross between them.
The optimum is the largest count any division can achieve.
For a large graph, finding the optimum can be difficult, so an approximation algorithm settles for a provably good fraction of it.
The release reports that the familiar guarantee of about eighty-seven point eight six percent cannot be improved in general by an efficient algorithm, under the assumption that P differs from NP.
Informally, that assumption says that being able to check a proposed answer efficiently does not always make it easy to find one. Here efficient means polynomial time: running time bounded by a fixed power of input size.
The release also reports the Unique Games Conjecture, a general statement about the difficulty of approximately satisfying certain systems of pairwise labeling rules. The guarantee concerns every possible input, not the fraction of edges in our example.
The main paper reports Unique Games hardness; a separate paper reports direct ordinary NP-hardness of beating the Goemans–Williamson ratio for simple unweighted Max-Cut. The graph shown is an illustrative cut with an exactly counted number of crossing edges. Its displayed best cut was checked exhaustively over its 4096 vertex partitions. It is not an instance used in the reduction.
The following descriptions are reproduced from OpenAI’s catalogue. They report the authors’ claims and retain the technical scope. The film does not independently certify their proofs.
Proves Khot's Unique Games Conjecture. Independent direct reductions also establish NP-hardness, on unweighted graphs, of approximation beyond the Goemans–Williamson ratio for Max-Cut, below factor two for Vertex Cover, and within any fixed constant factor for Min-UnCut and directed feedback vertex set. These direct proofs use established PCP and Label Cover hardness results.
A matrix is a rectangular table of numbers. Multiplying matrices combines two such tables using sums of products.
For square tables with n rows, the ordinary method uses roughly n cubed arithmetic operations. That exponent of three tells us how work grows: doubling n multiplies a pure cubic count by eight.
The release reports a multiplication exponent at most two point two five, for complex-number entries. Complex numbers are the two-coordinate numbers introduced in the zeta chapter. A pure two-point-two-five power grows by about four point seven six when n doubles.
The precise theorem allows any fixed positive extra amount in the exponent. That small allowance is usually called epsilon.
The result is asymptotic, meaning it describes growth for sufficiently large matrices. Hidden constants still matter, so it does not by itself establish a faster program on today's hardware.
The exponent ω over the complex numbers is reported to be ≤ 9/4. The displayed O(n^(2.25+ε)) statement is for every fixed positive ε. This is an arithmetic-operation asymptotic and does not report a measured implementation speedup. The illustrative doubling comparison is 2³=8 versus 2^2.25≈4.7568; it omits constants and ε. The displayed 2×2 numerical product is exact.
The following descriptions are reproduced from OpenAI’s catalogue. They report the authors’ claims and retain the technical scope. The film does not independently certify their proofs.
Proves over , giving arithmetic operations for square matrix multiplication. In characteristic zero, some inner dimension with permits rectangular multiplication. Further square bounds give outside finitely many positive characteristics and over every fixed field.
The algorithms package includes more than matrix multiplication and approximation barriers. One result removes randomness from computations using only logarithmic working memory. Logarithmic means that doubling the input needs only a constant number of extra memory cells. The claimed conversion keeps polynomial running time. This is a precise small-memory derandomization theorem, not a solution of every question about randomness in computation.
For multiplication of n-bit integers, another result improves on n times log n in the standard multitape computation model. The exponent improvement is extraordinarily small, two to the minus one hundred eighty-two. It is a theoretical improvement in long-run growth, not an advertised practical speedup.
A matching is a set of graph edges that share no endpoints. The release gives a randomized algorithm finding a largest matching in almost-linear time in the graph’s input size, with the stated success probability. The answer is exact when the algorithm succeeds.
Finally, polynomials can be factored using arithmetic modulo a prime, where remainders form a finite field. The package reports a deterministic polynomial-time algorithm for this problem, including repeated factors, with the prime supplied as part of the input.
The following descriptions are reproduced from OpenAI’s catalogue. They report the authors’ claims and retain the technical scope. The film does not independently certify their proofs.
Proves , resolving derandomization for bounded-error logarithmic-space computation. An effective compiler converts each randomized polynomial-time logarithmic-space machine deciding a language with one-sided or two-sided error into a deterministic logarithmic-space decider with explicit polynomial running-time bounds.
Multiplies two -bit integers exactly at every input length in deterministic worst-case time , with , on one fixed finite-alphabet Turing machine with finitely many one-dimensional tapes. This disproves the Schönhage–Strassen optimality conjecture in the ordinary multitape bit model.
Gives a randomized algorithm finding an exact maximum-cardinality matching in any simple undirected graph in word time, with success probability at least . The time bound holds on every computation path. The same guarantees apply to finding a spanning subgraph with prescribed admissible vertex degrees, or deciding that none exists.
Gives a uniform deterministic algorithm that completely factors every nonzero dense degree- polynomial over a prime field , including multiplicities, in bit complexity polynomial in . The prime is supplied in binary. No randomness, integer-factorization or primitive-root oracle, or GRH assumption is required.
A dynamical system describes how a state changes with time. A limit cycle is an isolated repeating motion, pictured as a closed orbit in a phase portrait. For planar polynomial motion rules of a fixed degree, the release reports a finite bound on how many limit cycles can occur. The bound depends only on the degree. This addresses the uniform boundedness part of Hilbert’s sixteenth problem.
Mixing means that observations far apart in time become statistically independent in the limiting sense. Pairwise independence does not by itself imply independence of several observations in ordinary probability. The Rokhlin result says that for an invertible mixing probability-preserving transformation, the dynamical mixing property does extend to every finite number of widely separated observations.
Chemical reaction networks supply a concrete application. Weak reversibility means that a directed reaction can be followed by a path of reactions returning to its starting combination of species. For finite mass-action networks of this kind with fixed positive rate constants, the package reports permanence. Every positive trajectory eventually stays within common positive lower and finite upper concentration bounds in its conservation class. This does not require convergence to a single equilibrium.
The following descriptions are reproduced from OpenAI’s catalogue. They report the authors’ claims and retain the technical scope. The film does not independently certify their proofs.
Resolves the uniform boundedness assertion in Hilbert's sixteenth problem: the number of isolated periodic orbits of a real planar polynomial vector field is bounded by a finite constant depending only on its degree. For classical quintic Liénard systems, the exact maximum is two limit cycles.
Proves that every invertible mixing probability-preserving transformation is mixing of all finite orders, resolving Rokhlin's multiple-mixing problem for a single transformation. Correlations among any finite collection of measurable sets converge to the product of their measures whenever all pairwise time separations diverge.
Proves the permanence conjecture for every finite weakly reversible mass-action system with fixed positive rate constants. Every positive stoichiometric compatibility class, even an unbounded one, has a common compact convex forward-invariant absorbing set. All positive trajectories in that class therefore eventually share positive lower and finite upper concentration bounds.
Seven, fourteen, twenty-one, twenty-eight: the gap is always seven. A sequence with constant spacing is called an arithmetic progression.
Now take an infinite set of positive integers. Replace each number by its reciprocal, meaning one divided by that number, and add them up.
A divergent sum means the running total grows without bound. Erdos conjectured that this condition forces progressions of every finite length. The release reports a proof.
A separate result concerns a sequence of plus and minus signs called the Liouville function. Factor a number into primes, counting repeats: twelve is two times two times three. An odd count gives minus one; an even count gives plus one.
The reported two-point Chowla result says that, for any fixed positive spacing, multiplying pairs of signs and averaging gives a value tending to zero. This is a precise kind of cancellation, not a claim that the signs are literally random.
The reciprocal-sum result concerns infinite sets of positive integers and progressions of every fixed finite length. The number line is an example, not a proof. The displayed Liouville values and their one-step shift are computed by counting prime factors with multiplicity. The Chowla statement concerns ordinary, not merely logarithmically averaged, two-point Liouville correlations along fixed nonproportional affine forms.
The following descriptions are reproduced from OpenAI’s catalogue. They report the authors’ claims and retain the technical scope. The film does not independently certify their proofs.
Proves Erdős's conjecture that every set of positive integers with divergent reciprocal sum contains arithmetic progressions of every finite length. Quantitatively, for each fixed , every subset of with no nonconstant -term progression has size at most , with positive constants depending only on .
Proves the ordinary two-point Chowla conjecture, with a bound for Liouville correlation sums along fixed nonproportional affine forms, where is absolute. More generally, proves the binary corrected Elliott conjecture for complex multiplicative functions bounded by one when one factor is uniformly nonpretentious against each fixed Dirichlet character times for .
Combinatorics often begins with pictures that are easy to state and hard to settle. Color every point of the plane so that points exactly one unit apart have different colors. The package reports that five colors cannot suffice. Since a seven-color construction is already known, the answer is narrowed to six or seven, not determined exactly.
A translational tile covers space using shifted copies without overlaps or gaps. A new three-dimensional tile is reported to cover space, but never in a fully periodic pattern. Another construction in nine dimensions disproves Borsuk’s proposal that every bounded set can be covered by ten smaller-diameter pieces. Diameter is the largest distance between two points of a set.
Graphs have surprises too. Sidorenko’s conjecture predicted that every bipartite pattern appears at least as often as in a random graph of the same edge density. Bipartite means the vertices split into two sets with all edges crossing between them. The reported counterexample has thirty-five vertices and sixty-six edges.
Finally, the crossing-number family determines the minimum possible number of edge crossings for every complete graph and every complete bipartite graph. A complete graph connects every pair of vertices. The result optimizes over drawings, rather than merely counting crossings in one chosen picture.
The following descriptions are reproduced from OpenAI’s catalogue. They report the authors’ claims and retain the technical scope. The film does not independently certify their proofs.
Constructs a finite translational tile in that tiles space but admits no fully periodic tiling, disproving the periodic tiling conjecture in the smallest possible lattice dimension. Its unit-cube thickening gives the same counterexample in , even with arbitrary real translation vectors.
Constructs a compact subset of that cannot be covered by ten sets of strictly smaller diameter, disproving Borsuk's covering assertion already in dimension nine. The example consists of rank-one orthogonal projectors onto lines in , with the Frobenius metric.
Proves that every five-coloring of the Euclidean plane has a monochromatic pair at distance one, with no restriction on the color classes. This advances the Hadwiger–Nelson problem: together with the classical seven-coloring, only six and seven remain possible chromatic numbers of the plane.
Disproves Sidorenko's conjecture with a connected bipartite pattern on vertices and edges that occurs less frequently than in a random graph of the same edge density. The same pattern disproves the forcing conjecture of Skokan and Thoma: matching its density and the edge density of a constant graphon need not force quasirandomness.
Resolves the Harary–Hill conjecture and Turán's brickyard problem in the Zarankiewicz formulation, determining the crossing numbers of every complete and complete bipartite graph. The result proves the optimality of the classical drawings among all plane drawings with continuous edge arcs.
With ordinary numbers, multiplication can undo multiplication: two times one half equals one in either order. Other algebraic systems need not behave like ordinary numbers.
A group describes reversible operations that can be combined, such as rotations.
A group algebra lets us add weighted combinations of group elements and multiply those combinations using the group's rules. Kaplansky's conjecture asked: in every group algebra, if A times B equals one, must B times A also equal one? The release reports a counterexample, an explicit construction showing the answer can be no.
It uses characteristic two, arithmetic in which one plus one equals zero.
A companion construction supplies a short way to see the failure: A times C equals zero, although C is not zero. If B times A were one, multiplying those identities would force C to be zero, a contradiction. The difficult part is constructing a group algebra where these relations really hold.
The September characteristic-two construction uses a finite field K, a finitely presented group G and elements a,b in K[G] with ab = 1 and ba ≠ 1. The October companion gives a finitely presented torsion-free counterexample over F₂. The listed Lean artifact is associated with the September construction. The animation shows the companion’s algebraic certificate: ab=1, ac=0 and c≠0 force ba≠1. It does not display the actual group presentation.
The following descriptions are reproduced from OpenAI’s catalogue. They report the authors’ claims and retain the technical scope. The film does not independently certify their proofs.
Constructs a finitely presented torsion-free nonsofic group whose group algebra over is not directly finite, disproving Kaplansky's conjecture even without torsion. Companion examples give injective nonsurjective cellular automata on all configurations, refuting Gottschalk's surjunctivity conjecture. Another counterexample is an integral group-ring matrix, invertible over the rational group ring, with Fuglede–Kadison determinant strictly between zero and one, disproving the unrestricted Determinant Conjecture.
Algebra studies systems in which we can add and multiply, including ones very different from ordinary numbers. An intersection multiplicity counts how geometric pieces meet, with weights for tangency and other complications. Serre’s result establishes the predicted strict positivity under the precise complementary-dimension and finite-intersection hypotheses, for the full class stated in the catalogue.
A zero divisor is a nonzero element that multiplies another nonzero element to zero. Ordinary real numbers have none, but matrices can. The package constructs them in a group algebra where Kaplansky’s conjecture predicted they could not occur. The group has no nonidentity element of finite order, so that familiar source of zero divisors is absent.
Other counterexamples concern resolutions: ways of analyzing an algebraic object through successive equations and relations. The little finitistic-dimension conjecture predicted a common upper bound on the lengths of the finite resolutions for a fixed finite-dimensional algebra. The reported example has finite lengths with no common bound. A separate family defeats several related structural conjectures, including Auslander–Reiten and Nakayama. Their exact statements are preserved in the catalogue.
The following descriptions are reproduced from OpenAI’s catalogue. They report the authors’ claims and retain the technical scope. The film does not independently certify their proofs.
Proves strict positivity of Serre's intersection multiplicity for nonzero finitely generated modules over any regular local ring, provided has finite length and . This resolves the positivity conjecture, including ramified mixed characteristic.
Constructs a finitely presented torsion-free group whose group algebra has nonzero zero divisors, disproving Kaplansky's zero-divisor conjecture. The group has a finite two-dimensional classifying space.
Constructs a finite-dimensional complex algebra whose finite-dimensional modules have unbounded finite projective dimensions. This disproves the little finitistic-dimension conjecture.
Constructs finite-dimensional algebras over a characteristic-two rational-function field that disprove the Auslander–Reiten and Gorenstein-projective conjectures, and Tachikawa's second conjecture. An associated endomorphism algebra also disproves the classical, generalized and strong Nakayama conjectures, the Auslander–Gorenstein conjecture, and the Wakamatsu tilting conjecture. The counterexamples persist under every extension of the base field.
Probability asks what becomes predictable when a system has many random parts. In percolation, connections are independently open or closed. An open cluster consists of points joined by open connections. The release reports that at the critical probability, no infinite cluster exists for the stated broad class of symmetric infinite graphs, and for both standard bond and site percolation on the three-dimensional cubic lattice.
Random planar maps are random networks drawn on a surface. One family reports a surface-to-tree transition in their large-size geometry as a model parameter crosses four. It also connects random walks on selected maps to motion on the limiting random surface.
Universality means that different microscopic rules can produce the same large-scale law. The Ising family studies two-state magnetic models. It reports that specified weak interactions and small random changes in interactions preserve the stated critical limits. The perceptron family, involving randomly constrained simple learning models, gives formulas for free energy, which balances energy and available arrangements, and results about the threshold where satisfying all constraints becomes impossible.
Random satisfiability asks when randomly generated true-or-false requirements stop being jointly satisfiable. A clause is a requirement such as A or not B or C; its size is the number of variables. For each fixed size at least three, the package reports a sharp limiting threshold and additional fluctuation results. It also reports computability of the three-SAT threshold. The paper explicitly credits Gaia Carenini with priority for threshold existence in concurrent work; its contributions include another proof and sharper variance and computability conclusions.
The following descriptions are reproduced from OpenAI’s catalogue. They report the authors’ claims and retain the technical scope. The film does not independently certify their proofs.
Critical Fortuin–Kasteleyn planar maps converge to Liouville quantum gravity spheres for and to the Brownian continuum random tree for , establishing the surface-to-tree geometric transition. For FK–Ising and spanning-tree-weighted maps, stationary random walks converge to Liouville Brownian motion on the limiting sphere. The FK–Ising spectral result also gives convergence of eigenvalues and heat traces, using the stated Brownian/LQG inputs.
Resolves the Benjamini–Schramm criticality conjecture for bond percolation on every infinite connected locally finite quasi-transitive graph with : at the critical probability, there is almost surely no infinite cluster. The family also establishes this conclusion for both nearest-neighbor bond and site percolation on .
Weak finite-range square-symmetric even multispin perturbations of the square-lattice Ising model preserve critical bulk spin and energy limits; weak square-symmetric contour interactions also yield chordal interface limits. With sufficiently weak iid bond disorder of any fixed bounded nondegenerate mean-zero law, critical spin interfaces converge to the same law in probability over environments.
Determines finite-temperature variational free energies for Gaussian Ising perceptrons with bounded Borel log-potentials and Gaussian spherical perceptrons with bounded continuous potentials, at every positive pattern density. A spherical extension treats bi-orthogonally invariant disorder with compact limiting singular-value distributions and no outliers. At margin , the quadratic-penalty spherical model has a sharp feasibility threshold and limiting gap and force laws, with system size, zero temperature, and critical density taken in that order.
For random -SAT with independent uniformly signed proper clauses sampled with replacement, proves finite positive limiting thresholds and hitting-time variance for every fixed , and computability of the 3-SAT threshold. We credit Gaia Carenini with priority for resolving the threshold-existence conjecture in her concurrent ECCC TR26-229, made public October 5, 2026; this family supplies another proof and the sharper variance and computability results.
Logic studies what formal rules can describe and what they force to be true. A mathematical theory can have many models, meaning structures that satisfy its rules. Categoricity means there is only one kind of model of a specified size, up to a structure-preserving match. The Shelah family reports that, for the stated class of theories, uniqueness at one sufficiently large size forces uniqueness at every larger size beyond a uniform threshold.
Turing degrees group computational problems by their power: two problems have the same degree if each can be solved using a solver for the other. Degrees are ordered by which problems can help solve which others. The release reports rigidity: no nontrivial relabeling preserves that entire order.
Another result turns any computably enumerable set into a polynomial equation whose auxiliary solution is unique for each member of the set, and absent for nonmembers. Computably enumerable means a program can list the members, without necessarily deciding every nonmember. Even with at most one solution promised, deciding existence can remain impossible.
Finally, the axiom of choice allows one selection from each set in an arbitrary collection of nonempty sets. A reported model of set theory satisfies a weaker principle comparing the sizes of sets and their images, while choice still fails. This is a relative consistency result, assuming the stated underlying set theory is consistent.
The following descriptions are reproduced from OpenAI’s catalogue. They report the authors’ claims and retain the technical scope. The film does not independently certify their proofs.
Proves Shelah's eventual categoricity conjecture in ZFC: for each bound on the Löwenheim–Skolem number, a uniform threshold makes categoricity of an abstract elementary class in one cardinal above that threshold imply categoricity throughout the same tail. Categoricity means uniqueness up to isomorphism at a given cardinality. Under the continuum hypothesis, a proposed specific Hanf threshold need not suffice.
Every order automorphism of the Turing degrees is the identity, resolving their rigidity problem. Thus no nontrivial relabeling of degrees preserves the ordering by relative computability.
Every recursively enumerable set of tuples of natural numbers has a Diophantine representation with exactly one auxiliary solution for each member and none for nonmembers. This proves the single-fold conjecture and hence the finite-fold conjecture. Diophantine solvability over the nonnegative integers remains undecidable even with an at-most-one-solution promise.
Assuming ZF is consistent, constructs a model in which every surjective image of a set injects into that set, yet the axiom of choice fails. Choice for ordinal-indexed families still holds. From any countable transitive model of ZFC, a separate construction gives a transitive symmetric extension with these properties and no new countable sequences of ground-model elements.
Recall that a group describes reversible moves. A finite presentation gives finitely many basic moves and finitely many relations, yet the resulting group can have infinitely many elements. One construction in the package is infinite even though every element eventually repeats after a power-of-two number of repetitions. It is also residually finite: distinct elements can be distinguished in suitable finite group models.
Groups can also have geometry. For a hyperbolic group, large triangles in its move graph are thin: each side stays near the other two. The boundary records directions toward infinity. Cannon’s conjecture says that if this boundary is a two-dimensional sphere, the group should act as geometric symmetries of hyperbolic three-space. The package reports that conclusion, including the stated compactness and finite-kernel conditions.
Amenability asks whether a group admits averaging compatible with its moves. For a finitely generated group, one can test whether large finite pieces can have arbitrarily small boundary relative to their size. The package reports that Thompson’s group F is nonamenable. This group is built from particular piecewise linear rearrangements of an interval.
The following descriptions are reproduced from OpenAI’s catalogue. They report the authors’ claims and retain the technical scope. The film does not independently certify their proofs.
Every word-hyperbolic group with boundary homeomorphic to admits a proper cocompact isometric action on hyperbolic three-space with finite kernel, proving Cannon's conjecture. Every torsion-free such group is therefore the fundamental group of a closed hyperbolic three-manifold.
Constructs an infinite finitely presented residually finite group whose elements all have finite -power order, answering the finitely presented Burnside problem negatively even in this class. The construction also yields an infinite-dimensional finitely presented nil associative -algebra and a finitely presented infinite-dimensional algebraic unitization, giving negative answers to the corresponding nilpotence and Kurosh finiteness questions.
Proves that Thompson's group , the group of dyadic piecewise linear homeomorphisms of the interval, is nonamenable, resolving its longstanding amenability problem.
Mathematical physics includes claims about spacetime itself. A trapped region is one where even outward-moving light fails to expand in the usual way. The Penrose inequality connects the area needed to enclose such a region with the total mass measured far away. In three spatial dimensions, the familiar geometric-units form compares mass with the square root of area divided by sixteen pi.
The release reports an enclosing-area spacetime version in every spatial dimension at least three, under the stated energy and decay assumptions, together with equality statements and charged extensions. These are mathematical statements about specified initial data, not measurements of astronomical black holes.
Strong cosmic censorship concerns whether the evolution determined by initial data can be extended beyond the region where prediction is unique. A separate result addresses generic data near a fixed rotating, subextremal Kerr configuration with two asymptotically flat ends. It rules out extensions in a specified weak regularity class. Both the neighborhood restriction and the exact meaning of an allowed extension are essential.
The following descriptions are reproduced from OpenAI’s catalogue. They report the authors’ claims and retain the technical scope. The film does not independently certify their proofs.
Proves the sharp enclosing-area spacetime Penrose inequality for smooth one-ended asymptotically flat initial data in every spatial dimension , under dominant energy, weak future trapping, positive enclosing area, and the stated decay assumptions. It bounds invariant ADM mass below using minimum enclosing area, with equality rigidity under additional horizon hypotheses. Charged upper-area bounds treat dyonic three-dimensional data; the higher-dimensional purely electric extension uses the matched neutral theorem.
Proves local strong cosmic censorship near each fixed rotating subextremal Kerr bridge. A dense subset of a weighted smooth neighborhood of smooth complete two-ended asymptotically flat vacuum data has full maximal globally hyperbolic developments with no future continuous nondegenerate extension whose weak connection is locally square-integrable. No symmetry is imposed; extensions need not satisfy the vacuum equations.
A ground state is a quantum system’s lowest-energy state. An energy gap separates it from the next allowed energy. A uniform gap stays positive as the system grows. The package reports such a gap for the spin-one antiferromagnetic Heisenberg chain on even periodic rings. Antiferromagnetic means neighboring spins favor opposite alignment.
A second gap result concerns the specified Laughlin model at one-third filling on a sphere, with stability under sufficiently weak scalar disorder. Filling describes the fraction of available magnetic states occupied by particles. The claim is for the stated idealized interaction, not every model of the fractional quantum Hall effect.
Entanglement entropy measures quantum entanglement between a region and the rest of a pure state. For the stated two-dimensional systems with a unique uniformly gapped ground state, an area law bounds this entropy by a constant times boundary length. The same family gives controlled approximations that can represent these states more economically.
Bose–Einstein condensation means that a macroscopic fraction of particles occupy the same quantum state. Another family reports condensation at a strictly positive temperature in a dilute three-dimensional hard-sphere gas, and a zero-temperature formula for quantum depletion, the fraction pushed out of that state by interactions.
The following descriptions are reproduced from OpenAI’s catalogue. They report the authors’ claims and retain the technical scope. The film does not independently certify their proofs.
Proves an entropy area law for unique ground states of finite-range Hamiltonians on arbitrary finite induced square-lattice domains, using only a uniform full-system spectral gap and bounds on the local interactions. On open squares, uniformly gapped nearest-neighbor ground states also admit projected entangled-pair state approximations with polynomial bond dimension and global vector error at most .
Proves Bose–Einstein condensation for the exact canonical Gibbs state of the three-dimensional hard-sphere gas: each fixed exclusion distance and sufficiently small fixed density admit a strictly positive temperature, independent of volume, with positive condensate fraction in the thermodynamic limit. At zero temperature, proves the Bogoliubov leading quantum-depletion law for hard spheres and fixed bounded nonnegative radial finite-range potentials of positive scattering length, taking the thermodynamic limit before the dilute limit.
Proves the spin-one Haldane gap conjecture for the pure antiferromagnetic Heisenberg chain on even periodic rings: the spectral gap stays uniformly positive as the chain grows. A companion establishes a gap for odd open chains with endpoint field and gives boundary-selected infinite-volume states with topological index .
Proves the fermionic Laughlin spectral-gap conjecture for the full interaction at filling on the round sphere. The unique ground state remains uniformly gapped under sufficiently weak bounded real scalar one-body potentials projected to the lowest Landau level. Both the gap and disorder threshold are uniform over all sufficiently large particle numbers and all normalized potential profiles.
Spin is a quantum property that makes particles behave, in some respects, like tiny magnets.
In a spin glass, some interactions favor alignment and others oppose it. These competing preferences can create many different arrangements.
Diluted means each spin interacts with only a finite average number of others.
Free energy balances energy against how many arrangements are available at a given temperature. The release reports the Mezard-Parisi formula for that quantity in a specified class of diluted spin glasses.
Now consider a different model, a quantum Heisenberg ferromagnet, where the interactions favor magnetic alignment.
Magnetization measures the resulting net magnetic moment. As the temperature rises slightly above zero, that magnetization decreases.
The reported Bloch law says the decrease is proportional to temperature to the three-halves power, with the exact multiplier determined by the interactions. The arrows are explanatory pictures of these quantum systems, not literal rotating magnets or a numerical simulation.
The Mézard–Parisi claim is restricted to diluted even-arity Ising models satisfying the Panchenko–Talagrand factorization and positivity assumptions, with first-moment integrability. The Bloch law concerns 3D quantum Heisenberg ferromagnets, fixed positive quantum spin, nonnegative symmetric finite-range couplings whose support generates Z³, with the stated order of limits. The drawn arrows and normalized low-temperature curve are illustrations, not computed solutions of these models.
The following descriptions are reproduced from OpenAI’s catalogue. They report the authors’ claims and retain the technical scope. The film does not independently certify their proofs.
Proves the Mézard–Parisi hierarchical cavity formula for Poisson-diluted even-arity Ising models satisfying the Panchenko–Talagrand factorization and positivity assumptions, with only first-moment integrability. The limiting free energy equals the infimum over finite-depth hierarchical trial laws. This includes the Viana–Bray model, symmetric diluted even-spin models, and weighted soft even- satisfiability.
Proves Bloch's law with its exact coefficient for three-dimensional quantum Heisenberg ferromagnets at every positive quantum spin, allowing nonnegative symmetric finite-range couplings whose support generates . The thermodynamic limit precedes the zero-field derivative and low-temperature limit. The family also proves spontaneous magnetization for nearest-neighbor models in every dimension and determines the first lattice correction for three-dimensional nearest-neighbor couplings.
A generator is a basic move from which more complicated moves are built.
In a free group, there are no extra relations between those moves: a move cancels its inverse, but otherwise different reduced sequences stay different.
Two generators and three generators give different groups, illustrated by these branching trees.
Now change the object we study. Think of a matrix as a rule that transforms a vector. An operator extends this idea to spaces that can have infinitely many coordinates. Bounded means an operator cannot stretch a vector by an arbitrarily large factor.
A von Neumann algebra is a collection of these operators that also contains the results of the relevant arithmetic operations and limits. Certain basic building blocks of these systems are called factors.
The release reports that the factors built from two and three free generators are isomorphic. Isomorphic means there is a one-to-one correspondence preserving their mathematical structure. So the groups differ, but the operator systems constructed from them are reported to be the same kind of object. Established results then extend this to a larger family of factors.
The reported isomorphism is between the tracial von Neumann algebras L(F₂) and L(F₃), not between the free groups F₂ and F₃. The consequence for all interpolated factors uses the classical free group factor alternative. Drawings are finite tree fragments illustrating the different generator counts.
The following descriptions are reproduced from OpenAI’s catalogue. They report the authors’ claims and retain the technical scope. The film does not independently certify their proofs.
Resolves the free group factor isomorphism problem: , and hence all interpolated free group factors, including , are isomorphic. Their common factor has fundamental group .
After the free-group-factor result, the operator-algebra catalogue continues in several directions. A representation realizes an abstract algebra as transformations of a vector space. Kadison’s similarity conjecture asks whether a bounded representation can be adjusted by a change of coordinates so that it also respects the adjoint operation, the analogue of a matrix’s conjugate transpose. The package reports that it can, under the exact complex-linear and identity-preserving assumptions.
The generator problem asks how many transformations are needed to build an entire operator algebra using its allowed operations and limits. The release reports that every type two-one factor with separable predual can be generated by one operator. The technical class is specified in the notes; the point is a single generating transformation for this broad class of infinite systems.
There are counterexamples too. An invariant subspace stays inside itself under a transformation. Hyperinvariant means it stays inside itself under every transformation commuting with that operator. The package constructs an operator with no nontrivial closed hyperinvariant subspace. This is a stronger preservation requirement than ordinary invariance, so the distinction matters.
The following descriptions are reproduced from OpenAI’s catalogue. They report the authors’ claims and retain the technical scope. The film does not independently certify their proofs.
Proves Kadison's similarity conjecture: every bounded complex-linear unital algebra homomorphism from a unital complex -algebra to operators on a Hilbert space becomes a -homomorphism after conjugation by a bounded invertible operator.
Constructs a nonzero norm-quasinilpotent operator on every infinite-dimensional separable complex Hilbert space with no nonzero proper closed subspace invariant under every commuting operator. The construction also gives operators with no nontrivial invariant projection in the hyperfinite type factor.
Proves that every type factor with separable predual is generated by a single operator, equivalently by two self-adjoint operators, resolving the generator problem. More strongly, for every irreducible inclusion of such factors, the unitaries with form a dense subset in the trace -norm topology.
Topology studies shapes through continuous changes. A manifold looks locally like ordinary Euclidean space. The Hilbert–Smith result says that a group acting continuously and faithfully on such a manifold must be a Lie group under the stated local-compactness and countability assumptions. A Lie group is a symmetry group with smoothly varying, finite-dimensional parameters.
Knot surgery removes a tube around a knot and glues it back in a specified way. The gluing slope records that choice. The cosmetic-surgery result says that two distinct slopes on a nontrivial knot in the three-sphere cannot give the same oriented three-dimensional manifold. Preserving orientation means respecting handedness.
Homotopy compares maps through continuous deformation, a weaker notion than a reversible change of coordinates. The package reports the homotopy hypothesis for the specified algebraic models of paths, paths between paths, and all higher levels. It also gives four-dimensional spaces that are homotopy equivalent but not homeomorphic, contradicting Borel rigidity in that setting. Matching the homotopy information does not force an actual homeomorphism.
The following descriptions are reproduced from OpenAI’s catalogue. They report the authors’ claims and retain the technical scope. The film does not independently certify their proofs.
Every locally compact second-countable Hausdorff group acting faithfully and jointly continuously on a connected finite-dimensional topological manifold is a Lie group. This proves the Hilbert–Smith conjecture in all finite dimensions, for Hausdorff second-countable manifolds without boundary.
Distinct Dehn surgery slopes on a nontrivial smooth knot in never produce orientation-preservingly homeomorphic manifolds, proving the purely cosmetic surgery conjecture. The statement includes the meridional slope.
Proves the Grothendieck homotopy hypothesis for -groupoids associated with every Grothendieck coherator in the Ara–Henry convention: these algebraic objects recover the homotopy theory of spaces.
Constructs closed connected aspherical topological four-manifolds that are homotopy equivalent but not homeomorphic, with a common word-hyperbolic fundamental group. This disproves even the homeomorphism-existence formulation of the Borel conjecture in dimension four.
Functional analysis treats functions, sequences, and vectors as points in a space. A norm assigns a length, and a Banach space is complete: sequences whose terms eventually become arbitrarily close have limits in the space. Tingley asks whether knowing the distances on the unit sphere determines the whole space. The reported answer is yes: every onto distance-preserving map between real Banach unit spheres extends uniquely to a linear distance-preserving map of the spaces.
The Crouzeix result bounds how strongly a polynomial of an operator can stretch vectors by testing that polynomial on a particular region of the complex plane called the numerical range. The sharp multiplier is two, also for the more general matrix-valued version in the paper. The numerical range consists of inner-product measurements of the operator along unit vectors.
A nonexpansive map never increases distances. A fixed point is a point it leaves unchanged. The package reports a fixed-point theorem for such maps on nonempty closed bounded convex subsets of real reflexive Banach spaces. Reflexivity is a compactness-related condition on the space; it is a substantive assumption, not a property of every Banach space.
The following descriptions are reproduced from OpenAI’s catalogue. They report the authors’ claims and retain the technical scope. The film does not independently certify their proofs.
Resolves Tingley's problem: every surjective isometry between the unit spheres of nonzero real Banach spaces extends uniquely to a surjective real-linear isometry of the whole spaces. No dimension restriction is imposed, so the metric geometry of the unit sphere determines the Banach space up to linear isometry.
Resolves the complete Crouzeix conjecture: for every bounded operator on a complex Hilbert space and every finite matrix-valued polynomial , one has , where is the numerical range. The constant is sharp, independent of the matrix size, and valid in infinite dimensions.
Resolves Kirk's reflexive-space fixed-point problem: every nonexpansive selfmap of a nonempty closed bounded convex subset of a real reflexive Banach space has a fixed point. The result uses the original norm, without assuming uniform convexity.
A surface metric specifies infinitesimal distances measured within the surface. An isometric immersion realizes those distances in an ambient space, while allowing the image to cross itself. The package reports that every closed smooth surface metric can be realized smoothly in four-dimensional Euclidean space. A separate construction gives a smooth metric that cannot be realized even locally in three-dimensional Euclidean space. The extra dimension changes what is possible.
Isoperimetry compares a region’s volume with the size of its boundary. In the flat plane, the circle encloses a given area using the least perimeter. The reported Cartan–Hadamard result extends the sharp comparison to every dimension for the stated complete, simply connected spaces with nonpositive curvature, along with generalized filling inequalities.
A geodesic is a path that is locally straight according to the surface’s own geometry. A closed geodesic returns to its starting point and direction. The package reports infinitely many geometrically distinct closed geodesics for every smooth metric on every sphere of dimension at least two, and on every closed three-manifold. Distinct means different paths, not merely repeated trips around one loop.
The following descriptions are reproduced from OpenAI’s catalogue. They report the authors’ claims and retain the technical scope. The film does not independently certify their proofs.
Every closed smooth Riemannian surface admits a smooth isometric immersion into , resolving the closed-surface form of the four-dimensional isometric-immersion problem. This includes nonorientable surfaces and metrics of arbitrary Gaussian curvature.
Constructs a smooth positive-definite metric on for which no neighborhood of the origin admits a smooth isometric immersion into . This answers the unrestricted smooth local isometric realization problem for surfaces negatively, even after shrinking the neighborhood.
Proves generalized Cartan–Hadamard isoperimetry in every dimension: in a complete simply connected manifold with sectional curvature at most , every finite-volume finite-perimeter set satisfies the sharp comparison with the equal-volume model ball. Bounded positive-volume equality regions for are Euclidean balls. Also proves sharp Euclidean filling bounds for compactly supported integral -cycles, , in arbitrary proper CAT spaces.
Proves that every smooth Riemannian metric on , , has infinitely many prime closed geodesics with pairwise distinct images. The same conclusion holds on every closed manifold admitting a finite smooth spherical cover and on every closed three-manifold, without orientability or nondegeneracy restrictions.
A plasma is matter containing freely moving charged particles. Their charges and motion produce electric and magnetic fields, which in turn affect the particles. The relativistic Vlasov-Maxwell equations model this feedback while respecting the speed limit of special relativity.
Rather than tracking each individual particle, they describe a distribution: how much particle density occupies different positions and momenta.
Momentum measures motion and, at low speeds, equals mass times velocity.
The mathematical question is whether an initially smooth distribution can develop a singularity: a breakdown of the smooth solution.
The release reports that, for its three-dimensional, one-species model, every allowed smooth starting state has a unique solution that stays smooth for all finite times.
One species means one kind of particle. The theorem allows large initial states, but still imposes precise conditions on the cloud and fields. Those conditions are listed in the guide. The moving dots illustrate the feedback; they do not prove the theorem.
The theorem is for the 3D one-species relativistic Vlasov–Maxwell system. Initial particle density is compactly supported; fields have finite energy and bounded derivatives of every order; smooth admissible data are required. There is no smallness or symmetry restriction. The trajectories in the video are analytic illustration paths, not an RVM solver.
The following descriptions are reproduced from OpenAI’s catalogue. They report the authors’ claims and retain the technical scope. The film does not independently certify their proofs.
Proves large-data global existence and uniqueness for the three-dimensional, one-species relativistic Vlasov–Maxwell system. Smooth admissible initial data may be arbitrary provided the particle density is compactly supported and the electromagnetic fields have finite energy and bounded derivatives of every order; the solution remains smooth on every finite time interval.
Differential equations relate a system to how it changes. The plasma theorem gave a global smooth evolution. Other equations in the package behave differently. For the stated periodic hard-sphere Boltzmann model of a gas, two different entropy solutions can start from the same density while both exactly conserve mass, momentum, and kinetic energy locally. These are solutions in a weaker regularity class, so this does not contradict uniqueness theorems for smooth solutions.
The hot-spots result concerns the slowest nonconstant mode of heat diffusion in a smooth bounded planar region with insulating boundary and no holes. Its hottest and coldest values must occur on the boundary. This is a statement about that eigenmode, not every arbitrary temperature pattern at every time.
De Giorgi concerns transition layers between two phases. The package reports that a bounded solution in eight dimensions, strictly increasing in one direction and satisfying the stated equation, must depend on just one linear coordinate. Its layers are therefore planar.
Finally, specially forced viscous Navier–Stokes flows are reported to perform universal computation. A prescribed particle reaches a target region exactly when the encoded computer program halts. The external forcing is part of the construction. This is a result about computational power and undecidability, not a resolution of the general unforced Navier–Stokes regularity problem.
The following descriptions are reproduced from OpenAI’s catalogue. They report the authors’ claims and retain the technical scope. The film does not independently certify their proofs.
Constructs two distinct global entropy solutions of the three-dimensional periodic hard-sphere Boltzmann equation from the same nonnegative initial density, with bounded velocity support and finite mass, energy and absolute entropy. Both are strongly continuous in and satisfy exact local conservation of mass, momentum and kinetic energy.
Proves a strict form of Burdzy's simply connected hot spots conjecture. On every smooth bounded simply connected planar domain, each nonzero eigenfunction for the first positive Neumann eigenvalue has no interior critical point, so all global extrema lie on the boundary. Eigenvalue multiplicity is allowed.
Proves De Giorgi's conjecture at its sharp dimension-eight endpoint: every entire solution of that is strictly increasing in one direction depends on only one linear coordinate. A stronger theorem classifies all stable entire solutions as constant wells or planar transitions, without an energy-growth assumption.
Constructs viscous incompressible flows starting from rest on a fixed flat three-dimensional domain that perform universal computation under smooth external forcing. A terminating compiler turns a Turing machine and input into a finite program for the force, so a designated particle reaches a fixed region exactly when the machine halts. The viscosity is fixed, positive and computable.
The package is a collection of positive theorems, sharper estimates, constructions, counterexamples, and connections between areas. Some claims settle a named conjecture. Others settle a restricted case, improve a bound, or depend on a companion result. Those distinctions are part of the mathematics.
Every subject has now appeared in the film. The companion page lists all three hundred seventy-two families and all seven hundred twenty-two manuscripts, with the original technical descriptions and source links. It identifies the families discussed here, includes definitions, and points to the formalization catalogue and available reasoning summaries. The next step is to examine the statements and proofs, with their assumptions visible.