ICM 2026 Invited Section Lecture - Will Sawin

Value of the Information & Strength of the Argument

The lecture provides substantial value by bridging classical number theory with probabilistic models and offering insights into recent research advances. The argumentation is logical and rigorous, starting from elementary observations (discriminants and quadratic forms) and systematically building up to sophisticated conjectures and theorems. The speaker clearly explains the reasoning behind Cohen-Lenstra heuristics, including the principle of weighting by automorphism groups, and then shows how known anomalies (e.g., the 2-part, real quadratic fields) naturally fit into a refined framework. His presentation of the non-abelian generalization is well-motivated, linking to the fundamental group of number fields and explaining both the challenges and the modifications needed (e.g., reflection theorem, self-duality). The final part on applications to sphere packings and the sum-product problem demonstrates the broader impact of these ideas, though these connections are presented more as illustrations than as fully developed results. Overall, the argumentation is compelling for an expert audience, with each step built on clear mathematical justification.

Scientific Rigor, Source Quality, Title Accuracy

The rigor of the lecture is high, as expected from an ICM invited talk. The speaker cites many foundational works, including those by Gauss, Heegner, Cohen and Lenstra, Golod and Shafarevich, Martinet, and Balog and Szemerédi, and also references recent collaborative results. However, no specific external sources are given in the video description, so the audience relies on the speaker’s presentation. The mathematical details are precise, and the speaker carefully distinguishes between conjectures, theorems, and open problems. The title is accurate but not descriptive of the content; it merely states that it is an invited lecture without mentioning the topic. This is acceptable for an ICM lecture but could be improved for a broader audience. The video structure is clear, with a logical progression from simple to complex ideas, and the speaker provides context and intuition throughout.

The title accurately indicates it is an invited lecture at ICM 2026 but is generic and does not mention the topic; the content matches the lecture format.

8/10

The speaker is a leading researcher in number theory, delivering an invited lecture at the International Congress of Mathematicians. The content is technically rigorous and well-structured, with clear reasoning and references to established work.

Markers derived by PSI from the transcript: the creator did not define chapters.

• 00:00Introduction: quadratic discriminants and equivalence classes of forms, leading to class groups.
• 05:00Cohen-Lenstra conjectures: probabilistic model for odd P-parts of class groups; principle of weighting by automorphisms.
• 13:20Failures and modifications: P=2, real quadratic fields, genus theory, Erdős-Kac law, role of Galois actions and roots of unity.
• 23:20Non-abelian Cohen-Lenstra: fundamental group of number fields, Golod-Shafarevich theorem, applications to sphere packings.
• 33:20Connection to sum-product problem: constructions from arithmetic and geometric progressions, recent improvements.

• Cohen–Lenstra heuristics — The speaker discusses the Cohen-Lenstra program and its probabilistic conjectures for class groups.
• Golod–Shafarevich theorem — Referenced when discussing infinite fundamental groups and their group-theoretic properties.
• Class group — The main object of study: ideal class groups of number fields.