VibeMathedMath problems solved with AI

Kalai's conjecture for tight trees

Kalai proposed in 1984, in a form recorded by Frankl and Füredi, a hypergraph generalisation of the Erdős-Sós conjecture: for a suitable notion of a tight tree TT with tt edges in an rr-uniform hypergraph, every rr-graph on nn vertices with more than t1r(nr1)\tfrac{t-1}{r}\binom{n}{r-1} edges should contain TT. The case r=2r=2 is the Erdős-Sós conjecture, proved in 2026. Does the conjecture hold for all rr?

Result
Proved
Status
Resolved
AI contribution
AI-discovered
Method
Argument
Field
Extremal hypergraph theory
Posed by
Gil Kalai (1984), recorded by Péter Frankl and Zoltán Füredi; the $r=2$ case is the Erdős-Sós conjecture
Year posed
1984
Years open
42y
Solved
2026-09-07
Model
GPT-6 Astra
Vendor
OpenAI
Collaborators
Dhruv Mubayi, Jacques Verstraëte
Verification
Unreviewed
Publication
Preprint
Significance
45 / 100
Disclosed cost
Wikipedia
No dedicated article

What the AI did

From the paper: the proof was found by GPT-6 Astra, extending its own method of proof for the Erdős-Sós conjecture to the hypergraph setting. The authors supplied the framing and the write-up.

Verification

Checked here on 22 September 2026 against arXiv:2609.08012: the introduction states that Kalai proposed the hypergraph generalisation in 1984, recorded by Frankl and Füredi, that the r=2r=2 case is the Erdős-Sós conjecture, and that the conjecture is proved here via a shadow bound whose equivalence to Kalai's conjecture was established by Füredi, Jiang, Kostochka and the authors. The mathematics was not checked here; fifteen days old, no referee. Same authors and same model as the digraph entry three days later.

Sources

Changelog1 change

Discussion