Soohak is a new 439-problem mathematician-authored benchmark showing frontier LLMs reach only 30% on research math and fail to exceed 50% on refusing ill-posed questions.
First Proof.arXiv preprint
8 Pith papers cite this work. Polarity classification is still indexing.
8
Pith papers citing it
years
2026 8representative citing papers
An interactive AI workbench for mathematicians achieves 48% on FrontierMath Tier 4 and helped solve open problems in early tests.
FlowBoost finds the Hermite pair as the unique equality case for the p=2 finite free Stam inequality, conjectures that the singular values of the coupling matrix E_n are 2^{-k/2} independent of n, and reveals a phase transition at the critical exponent p*=2 with bifurcating extremals for p<2.
k-server-bench formulates potential-function discovery for the k-server conjecture as a code-based inequality-satisfaction task; current agents fully solve the resolved k=3 case and reduce violations on the open k=4 case.
An AI framework combining informal reasoning and formal verification resolves an open commutative algebra problem and produces a Lean 4-checked proof with minimal human input.
pAI/MSc is a customizable multi-agent system that reduces human steering by orders of magnitude when turning a hypothesis into a literature-grounded, mathematically established, experimentally supported manuscript draft in ML theory.
Forage V2 enables agent organizations to grow knowledge from 0 to 54 entries over runs and transfer it so weaker models nearly match stronger ones in coverage, cost, and speed on open-world tasks.
AI agents exploring Platonic mathematical structures via proof hypergraphs may reveal the overall architecture of formal mathematics and what makes parts of it human-accessible.
citing papers explorer
-
Soohak: A Mathematician-Curated Benchmark for Evaluating Research-level Math Capabilities of LLMs
Soohak is a new 439-problem mathematician-authored benchmark showing frontier LLMs reach only 30% on research math and fail to exceed 50% on refusing ill-posed questions.
-
AI co-mathematician: Accelerating mathematicians with agentic AI
An interactive AI workbench for mathematicians achieves 48% on FrontierMath Tier 4 and helped solve open problems in early tests.
-
FlowBoost Reveals Phase Transitions and Spectral Structure in Finite Free Information Inequalities
FlowBoost finds the Hermite pair as the unique equality case for the p=2 finite free Stam inequality, conjectures that the singular values of the coupling matrix E_n are 2^{-k/2} independent of n, and reveals a phase transition at the critical exponent p*=2 with bifurcating extremals for p<2.
-
$k$-server-bench: Automating Potential Discovery for the $k$-Server Conjecture
k-server-bench formulates potential-function discovery for the k-server conjecture as a code-based inequality-satisfaction task; current agents fully solve the resolved k=3 case and reduce violations on the open k=4 case.
-
Automated Conjecture Resolution with Formal Verification
An AI framework combining informal reasoning and formal verification resolves an open commutative algebra problem and produces a Lean 4-checked proof with minimal human input.
-
pAI/MSc: ML Theory Research with Humans on the Loop
pAI/MSc is a customizable multi-agent system that reduces human steering by orders of magnitude when turning a hypothesis into a literature-grounded, mathematically established, experimentally supported manuscript draft in ML theory.
-
Forage V2: Knowledge Evolution and Transfer in Autonomous Agent Organizations
Forage V2 enables agent organizations to grow knowledge from 0 to 54 entries over runs and transfer it so weaker models nearly match stronger ones in coverage, cost, and speed on open-world tasks.
-
Artificial Intelligence and the Structure of Mathematics
AI agents exploring Platonic mathematical structures via proof hypergraphs may reveal the overall architecture of formal mathematics and what makes parts of it human-accessible.