What Happened
- Researchers have formally verified Maryna Viazovska's Fields Medal-winning proof of the sphere packing problem in 8 dimensions using the Lean theorem prover — a computer program that mechanically checks mathematical proofs step by step.
- The formalisation project, initiated at EPFL (École Polytechnique Fédérale de Lausanne) in March 2024 by Viazovska herself and Sidharth Hariharan, proves that Viazovska's 2016 result is computationally verifiable — not merely socially trusted.
- Mathematicians argue that "auto-formalisation" — the machine-verification of complex mathematical proofs — will eventually make mathematical correctness less dependent on trust among expert peers and more on explicit, verifiable machine-checkable reasoning.
- The sphere packing problem — how densely identical spheres can be packed in n-dimensional space — was solved by Viazovska in 8 dimensions (2016) and 24 dimensions (collaborative, 2016), earning her the Fields Medal in 2022.
- The significance extends beyond mathematics: formal verification techniques underpin software correctness proofs, cryptographic security proofs, and AI system safety verification.
Static Topic Bridges
Fields Medal — The Nobel Prize of Mathematics
The Fields Medal is the most prestigious award in mathematics, awarded every four years to two to four mathematicians under the age of 40 at the International Congress of Mathematicians (ICM). It was established by Canadian mathematician John Charles Fields in 1936.
- Awarded since: 1936 (ICM, Oslo); medal carries a cash prize of Canadian $15,000
- Frequency: Every 4 years, at the International Congress of Mathematicians
- Age restriction: Recipients must be under 40 years old
- 2022 Fields Medals awarded at ICM Helsinki (virtual due to Russia-Ukraine context): Maryna Viazovska, Hugo Duminil-Copin, James Maynard, June Huh
- Maryna Viazovska: Ukrainian mathematician at EPFL; second woman ever to win (after Maryam Mirzakhani, 2014)
- Indian connection: No Indian national has won the Fields Medal; however, Manjul Bhargava (Indian-origin Canadian-American) won in 2014
Connection to this news: The machine verification of Viazovska's Fields Medal-winning proof is significant because Fields Medal proofs are inherently complex and peer-reviewed by a small community of specialists — the Lean formalisation extends the verification to a mechanically checkable system that any trained programmer can audit.
Sphere Packing Problem — From Kepler to Viazovska
The sphere packing problem asks: what is the densest way to pack identical spheres in n-dimensional space? In 3 dimensions, this is the "Kepler conjecture" — that face-centred cubic (FCC) packing with density π/(3√2) ≈ 74% is optimal. Thomas Hales proved the 3D case in 1998, but the proof was so complex (250 pages + 3 GB of computer code) that peer reviewers took years to verify it — prompting renewed interest in formal verification.
- Kepler conjecture (1611): Proposed by Johannes Kepler; the optimal 3D packing density is π/(3√2) ≈ 74.048%
- Hales proof (1998): 3D Kepler conjecture proved by Thomas Hales; formally verified in Lean in 2015 (Flyspeck project, 20+ years later)
- Viazovska's 8D proof (2016): Proved E8 lattice packing achieves optimal density of π4/384 in 8 dimensions; solved a problem open since the 1970s — and the proof was described as "stunningly simple" compared to the 3D case
- 24D proof (2016): Viazovska and collaborators (Henry Cohn, Abhinav Kumar, David de Laat, Danylo Radchenko) simultaneously proved optimality of Leech lattice packing in 24 dimensions
- Dimensions 4-7, 9-23, 25+: Optimal packing density remains unknown
Connection to this news: The Lean formalisation of Viazovska's 8D result means that, for the first time, a Fields Medal-winning proof exists in a form that can be checked by a computer program — extending the legacy of the Flyspeck project (Hales' 3D proof) to higher dimensions.
Lean Theorem Prover — Formal Verification in Mathematics and Computing
Lean is an open-source interactive theorem prover and functional programming language developed at Microsoft Research by Leonardo de Moura. It allows mathematicians to write formal proofs that are mechanically checked for correctness by the computer. A companion project, Mathlib, provides a growing library of formalised mathematics in Lean.
- Lean developed at: Microsoft Research; current version Lean 4 released 2023
- Mathlib library: Community-maintained library of formalised mathematics in Lean 4; covers undergraduate and graduate-level mathematics
- Auto-formalisation: The (still-emerging) capability to automatically translate informal mathematical text into Lean's formal language — currently requiring significant human assistance
- Other proof assistants: Coq (used for CompCert formally verified C compiler), Isabelle (used in banking and aerospace), Agda, HOL4
- Applications beyond pure math: Formal verification in software engineering (proving code correctness), cryptography (proving security of encryption schemes), hardware design (Intel and AMD use formal methods for chip verification)
Connection to this news: The Lean formalisation of Viazovska's proof is a landmark in the "digitalisation of mathematics." Once a proof exists in Lean, it is eternally verifiable — any claim that the proof is incorrect must itself be a formally verifiable counterargument. This shifts mathematics from a social/trust-based process (expert peer review) to a computational/algorithmic one.
Key Facts & Data
- Viazovska's 8D sphere packing paper: published March 2016; arXiv preprint solved a problem open for decades
- Optimal packing density in 8D: π4/384 (achieved by E8 lattice)
- Fields Medal awarded to Viazovska: July 2022, ICM (International Congress of Mathematicians)
- Viazovska: Second woman to win Fields Medal (after Maryam Mirzakhani, 2014)
- Lean formalisation project started: March 2024, at EPFL (Lausanne, Switzerland)
- Project team: Christopher Birkbeck, Sidharth Hariharan, Bhavik Mehta, Seewoo Lee (initially kickstarted by Viazovska and Hariharan)
- Fields Medal: Awarded since 1936; every 4 years; age limit under 40; prize C$15,000
- Kepler conjecture (3D): Optimal packing density ≈ 74.048%; proved by Thomas Hales (1998); formalised in Lean by 2015 (Flyspeck project)
- Open dimensions: Optimal sphere packing still unknown in dimensions 4–7, 9–23, 25 and above