VibeMathedMath problems solved by AI

The Matrix Spencer Conjecture for Finite Groups

The group version of the Matrix Spencer conjecture holds: for every finite group GG there are signs ε{±1}G\varepsilon \in \{\pm 1\}^G with gGεgρ(g)CG\left\|\sum_{g \in G} \varepsilon_g \rho(g)\right\| \le C\sqrt{|G|}, where ρ\rho is the left regular representation and CC is universal.

Result
Proved(see note)
Status
Resolved
AI contribution
AI co-developed
Method
Argument
Field
Discrepancy theory
Posed by
Nikhil Bansal, Haotian Jiang, Raghu Meka
Year posed
2022
Years open
4y
Solved
2026-06-10
Model
ChatGPT Pro 5.5, Claude Opus 4.7 and 4.8
Vendor
OpenAI / Anthropic
Collaborators
Afonso S. Bandeira, Helmut Bolcskei
Verification
Unreviewed
Publication
Preprint
Significance
30 / 100
Disclosed cost
Wikipedia
No dedicated article

What was actually shown

the group case; the full Matrix Spencer conjecture remains open

What the AI did

The acknowledgement says modern AI tools were used throughout, and singles out one of the key arguments as the one that broke the problem open, crediting it to that use.

Verification

arXiv preprint; not yet peer-reviewed.

Sources

arXiv:2606.12181 - Matrix Discrepancy for Representations of Finite Groups

Discussion