MinF-2F-test
Emerging7papers using it
2025first seen
The 'MinF2F-test' dataset/benchmark is used to evaluate the performance of reinforcement learning algorithms in Automated Theorem Proving by providing a set of tasks that require the construction of machine-verifiable proofs.
Papers using MinF-2F-test (7)
- StepFun-Prover Preview: Let's Think and Verify Step by StepSTP: Self-play LLM Theorem Provers with Iterative Conjecturing and
ProvingDeepSeek-Prover-V2: Advancing Formal Mathematical Reasoning via Reinforcement Learning for Subgoal DecompositionRewarding the Unlikely: Lifting GRPO Beyond Distribution SharpeningGAR: Generative Adversarial Reinforcement Learning for Formal Theorem ProvingGDEPO: Group Dual-dynamic and Equal-right Advantage Policy Optimization with Enhanced Training Data Utilization for Sample-Constrained Reinforcement LearningPhysProver: Advancing Automatic Theorem Proving for Physics