Research·2 min read
By BitsMindsSource: Google DeepMind

Google DeepMind's AI Co-Mathematician Cracks a 60-Year-Old Group Theory Problem

DeepMind's new hierarchical research agent has resolved Problem 21.10 from the Kourovka Notebook, a group theory question that had stood unsolved since 1965, and topped FrontierMath Tier 4 at 48% accuracy — but the team is also publicly flagging a 'reviewer-pleasing bias' that nearly let a flawed proof slip through.

Google DeepMind's AI Co-Mathematician Cracks a 60-Year-Old Group Theory Problem
Share:

Google DeepMind on Tuesday published a paper describing AI Co-Mathematician, a hierarchical research agent built on Gemini 3.1 Pro that the company says has resolved Problem 21.10 from the Kourovka Notebook — a group theory question that had remained open since it was first posed in 1965. The proof was developed in collaboration with Oxford mathematician Marc Lackenby, who reviewed the system's work and provided the corrections needed to close the final gap. The paper is posted as arXiv:2605.06651.

On FrontierMath Tier 4, the hardest tier of a benchmark whose problems typically take professional mathematicians weeks to solve, AI Co-Mathematician scored 48%, well ahead of the 19% achieved by the base Gemini 3.1 Pro and the 39.6% reported for GPT-5.5 Pro. The lift, DeepMind argues, does not come from a larger underlying model so much as from a stacked architecture: a proposer agent generates candidate proofs, a reviewer agent critiques them in detail, and a separate planner reroutes the search when a chain of reasoning is judged unsound. The system also keeps explicit lemma libraries that persist across runs so partial progress is not thrown away between sessions.

The most striking section of the paper is not the headline result but an unusually candid failure analysis. The team reports what they call 'reviewer-pleasing bias' — a tendency for review agents to ratify proofs that match the reviewer's expectations of how a solution should look, even when those proofs contain subtle gaps. The Kourovka 21.10 proof itself contained one such flaw in its first internal draft; the system's reviewer agent did not catch it, and only Lackenby's outside read surfaced the problem. DeepMind frames this as a load-bearing limitation for any agentic system claiming superhuman reasoning: a reviewer that learned its standards from human-written proofs inherits the same blind spots.

The implications run well beyond pure mathematics. AI Co-Mathematician is the clearest demonstration to date that hierarchical agents — with explicit roles for generation, critique and planning — can outperform a single large model on tasks that demand multi-week, multi-step expert reasoning. That blueprint is already being copied by labs building agents for theorem-proving codebases, drug design and theoretical physics, and the reviewer-bias finding is likely to land as a cautionary note in each of those domains long before the Kourovka proof itself is formally verified.

Want AI news before everyone else?

The morning's most important AI stories, straight to your inbox. No fluff.

Related Articles

GPT-6 Astra and a lattice definition of the Standard Model Editorial illustration on a midnight-blue field: a silver isometric lattice of rods and nodes, standing for spacetime cut into a grid, with a silver coin engraved with the OpenAI logo at its centre and one square loop on its front face glowing gold. The type column reads OpenAI Science, GPT-6 Astra, the Standard Model on a lattice. The object is an editorial metaphor, not a diagram from the paper, and the result it refers to is a candidate construction that has not been peer reviewed. OPENAI SCIENCE GPT-6 Astra THE STANDARD MODEL ON A LATTICE BITSMINDS.COM
Research

GPT-6 Astra Helps Put the Standard Model on a Lattice

Google DeepMind Gemini Robotics 2.5 Editorial illustration on a blue-to-coral Gemini gradient: a white two-finger robot gripper on a graphite wrist pinches a porcelain puck carrying the Gemini spark. The type column reads Google DeepMind, Gemini Robotics 2.5, vision, language, action. An editorial metaphor, not a Google product. GOOGLE DEEPMIND Gemini Robotics 2.5 VISION · LANGUAGE · ACTION BITSMINDS.COM
Research

Gemini Robotics 2.5 Hits 96% on a Mercedes-Benz Line

Google DeepMind AlphaProof Code verifies Signal Editorial illustration on a deep navy field: a machined steel padlock tilted toward the viewer, its navy enamel face carrying the DeepMind logo in white, with a small green plaque reading PROVED. The type column reads Google DeepMind, AlphaProof Code, Signal verified in Lean. An editorial metaphor, not a Google or Signal product. GOOGLE DEEPMIND AlphaProof Code SIGNAL · VERIFIED IN LEAN BITSMINDS.COM PROVED
Research

AlphaProof Code Verifies Signal Down to Machine Code