Net Good IndexSubmit a correction

Benefit Ledger · provisional · Mathematics

Liu, Luo, and Xie post a complete proof of the Tu–Deng conjecture

On 30 July 2026 Renzhang Liu, Hengyi Luo, and Tianyuan Xie posted a complete proof of the Tu–Deng combinatorial conjecture on Hamming weights modulo 2^k−1 (arXiv:2608.05187). A compiled dataset records ChatGPT 5.6 Pro assistance; the arXiv abstract does not name a model or split of labor.

30 Jul 2026Tier 2 NotableMethodology 0.1

Current score

+0.13

3 base · Notable (tier 2 of 5, 3 pts)
× 0.3500 attribution · Material acceleration
× 0.3500 evidence · Technical disclosure or preprint
× 0.5000 realization · Experimentally validated
× 0.7000 durability
Event-level product before credit split: 0.13

A complete proof of a named coding-combinatorics conjecture is notable (tier 2). AI share is unspecified in the abstract (0.35). Evidence is a preprint (0.35). Realization is a public complete argument (0.50).

What happened

The Tu–Deng conjecture bounds the number of pairs (a,b) with a+b ≡ t (mod 2^k−1) and wt(a)+wt(b)<k. The preprint gives a complete argument via cyclic-carry enumerators. Because the public abstract does not specify the AI role, attribution is conservative and the ChatGPT 5.6 Pro identification is treated as reported, not as a named-release certainty. Partial Lean formalization is reported in secondary problem indexes.

Model attribution

GPT-5.6

Reported as an assistant on the Tu–Deng proof; the arXiv abstract does not name the model or the split of labor.

Secondary compilation credits ChatGPT 5.6 Pro; primary abstract is silent, so attribution stays modest and identity is the 5.6 Pro SKU rather than Sol.

Attribution 0.3500 · Credit share 100% · OpenAI

Claims

  • Liu, Luo, and Xie posted a complete proof of the Tu–Deng conjecture.

    outcome · supported

  • GPT-5.6 Pro was the primary author of the proof.

    attribution · disputed

Sources

primary sources

Secondary domains: Computer Science

Revision history

  • 13 Sep 2026 · 0.00 0.13

    Imported events under methodology 0.1.