analiza aiqbench
This study challenges the necessity of specialized tooling for formal theorem proving by demonstrating that a minimal agent using only standard shell commands can outperform complex baselines on the Putnam 2025 benchmark. The findings suggest that improvements in base model capabilities may be sufficient to drive significant gains, potentially simplifying the architecture of future AI-driven mathematical verification systems.
Do Theorem Provers Need Theorem-Proving Agents? mini-Lean-agent: A Coding-Agent Baseline for Lean
arXiv:2610.17331v1 Announce Type: new Abstract: General-purpose coding agents use standard shell and file interactions, yet formal theorem-proving agents often rely on specialized, ad-hoc tooling. This makes it difficult to determine which components of theorem-proving agents provide benchmark gains: an improved base model, a specialized harness, a specific tool, or a combination of any of them. We introduce mini-Lean-agent, a deliberately minimal agent for Lean tasks whose only model-visible tool is bash. The model must use bash to inspect and edit the repository, search local libraries, invoke Lean, and manage its own intermediate work, while a separate trusted evaluator verifies the final artifact. We compare mini-Lean-agent against AxProverBase and Numina-Lean-Agent on the Putnam 2025 benchmark. Experiments with Gemini 3.8 Flash as the base model show that mini-Lean-agent solves 7/12 problems at \$2 per problem limit, compared with AxProverBase's 4/12. With Astra as the base model, mini-Lean-agent solves 12/12 problems at an average cost of \$1.28 per problem. This suggests that a generic coding agent with a minimal harness can already achieve strong performance on Lean benchmarks. Our results motivate using mini-Lean-agent as a standard baseline when evaluating theorem-proving infrastructure.
Title:Do Theorem Provers Need Theorem-Proving Agents? mini-Lean-agent: A Coding-Agent Baseline for Lean
Abstract:General-purpose coding agents use standard shell and file interactions, yet formal theorem-proving agents often rely on specialized, ad-hoc tooling. This makes it difficult to determine which components of theorem-proving agents provide benchmark gains: an improved base model, a specialized harness, a specific tool, or a combination of any of them. We introduce mini-Lean-agent, a deliberately minimal agent for Lean tasks whose only model-visible tool is bash. The model must use bash to inspect and edit the repository, search local libraries, invoke Lean, and manage its own intermediate work, while a separate trusted evaluator verifies the final artifact. We compare mini-Lean-agent against AxProverBase and Numina-Lean-Agent on the Putnam 2025 benchmark. Experiments with Gemini 3.8 Flash as the base model show that mini-Lean-agent solves 7/12 problems at \$2 per problem limit, compared with AxProverBase's 4/12. With Astra as the base model, mini-Lean-agent solves 12/12 problems at an average cost of \$1.28 per problem. This suggests that a generic coding agent with a minimal harness can already achieve strong performance on Lean benchmarks. Our results motivate using mini-Lean-agent as a standard baseline when evaluating theorem-proving infrastructure.
Artificial Intelligence (cs.AI); Logic in Computer Science (cs.LO)
I.2.3; I.2.8; F.4.1
arXiv:2610.17331 [cs.AI]
(or arXiv:2610.17331v1 [cs.AI] for this version)
https://doi.org/10.48550/arXiv.2610.17331 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 297 cuvinte
Subiecte conexe
Can you hear the shape of a Lean soundness bug? A wager.
LessWrong · 10 sept. 2026

A critical look at Lean4
LessWrong · 9 oct. 2026

Deconfusing Corrigibility
AI Alignment Forum · 17 oct. 2026
Cost-Efficient Theorem Proving via Agent Orchestration in Program Verification
arXiv · 8 oct. 2026