VibeMathedMath problems solved by AI

A Generalization of Boppana's Entropy Inequality

A generalization of Boppana's entropy inequality, of the kind used in union-closed-sets arguments, proved and formalized: the sharp form with the extremal constant characterized via the unique positive solution of an explicit equation.

Result
Proved
Status
Resolved
AI contribution
AI co-developed
Method
Argument
Field
Entropy inequalities
Posed by
Year posed
Years open
Solved
2026-01-27
Model
GPT-5.2 Pro, Harmonic Aristotle, Gemini 3 Pro
Vendor
OpenAI, Harmonic, Google DeepMind
Collaborators
Boon Suan Ho
Verification
Lean-checked, statement unaudited
Publication
Preprint
Significance
8 / 100
Disclosed cost
Wikipedia
No dedicated article

What the AI did

"Key steps in some proofs were generated with the assistance of GPT-5.2 pro. The result has also been formalized in Lean 4 using Harmonic Aristotle and Gemini 3 Pro Preview" - with the formalization public.

Verification

Formalized in Lean 4 (Aristotle plus Gemini); code public on GitHub. No independent review. Tier: the formalization was produced by the assisting systems and checked by the author; no independent statement audit.

Sources

arXiv

Discussion