MATH-500

Dataset Card for MATH-500 This dataset contains a subset of 500 problems from the MATH benchmark that OpenAI created in their Let's Verify Step by Step paper. See their GitHub repo for the source file: https://github.com/openai/prm800k/tree/main?tab=readme-ov-file#math-splits

Papers using MATH-500 (61)

Reasoning or Memorization? Unreliable Results of Reinforcement Learning Due to Data Contamination2025 · 68 citesStable Reinforcement Learning for Efficient Reasoning2025 · 25 citesNemotron-CrossThink: Scaling Self-Learning beyond Math Reasoning2025 · 24 citesSqueeze the Soaked Sponge: Efficient Off-policy Reinforcement Finetuning for Large Language Model2025 · 10 citesKimi k1.5: Scaling Reinforcement Learning with LLMs2025 · 11 citesQeRL: Beyond Efficiency -- Quantization-enhanced Reinforcement Learning for LLMs2025 · 4 citesHow Much Online RL is Enough? Informative Rollouts for Offline Preference Optimization in RLVR2026RL with Learnable Textual Feedback: A Bilevel Approach2026MOTIF: Modular Thinking via Reinforcement Fine-tuning in LLMs2025 · 2 citesReinforcement Learning for Reasoning in Large Language Models with One Training Example2025 · 1 citesDiscrete Tilt Matching2026R$^2$PO: Decoupling Training Trajectories from Inference Responses for LLM Reasoning2026DianJin-R1: Evaluating and Enhancing Financial Reasoning in Large Language Models2025 · 1 citesProcess Reward Models That Think2025 · 1 citesOpen-Reasoner-Zero: An Open Source Approach to Scaling Up Reinforcement Learning on the Base Model2025 · 1 citesSLIM-RL: Risk-Budgeted Random-Masking RL for Diffusion LLMs Without Trajectory Slicing2026CATPO: Critique-Augmented Tree Policy Optimization2026Maximizing Rollout Informativeness under a Fixed Budget: A Submodular View of Tree Search for Tool-Use Agentic Reinforcement Learning2026Gradient Starvation in Binary-Reward GRPO: Why Group-Mean Centering Fails and Why the Simplest Fix Works2026PopuLoRA: Co-Evolving LLM Populations for Reasoning Self-Play2026How Much Thinking is Enough? Quantifying and Understanding Redundancy in LLM Reasoning2026ESPO: Early-Stopping Proximal Policy Optimization2026Reasoning with Sampling: Cutting at Decision Points2026LLM Reasoning with Process Rewards for Outcome-Guided Steps2026PerMix-RLVR: Preserving Persona Expressivity under Verifiable-Reward Alignment2026Sampling for Quality: Training-Free Reward-Guided LLM Decoding via Sequential Monte Carlo2026Tool Verification for Test-Time Reinforcement Learning2026SortedRL: Accelerating RL Training for LLMs through Online Length-Aware Scheduling2026Think Dense, Not Long: Dynamic Decoupled Conditional Advantage for Efficient Reasoning2026PrAg-PO: Prompt Augmented Policy Optimization for Robust and Diverse Mathematical Reasoning2026Boosting LLM Reasoning via Human-Inspired Reward Shaping2026Beyond Correctness: Learning Robust Reasoning via Transfer2026Latent Poincar\'e Shaping for Agentic Reinforcement Learning2026Long Chain-of-Thought Compression via Fine-Grained Group Policy Optimization2026Prioritize the Process, Not Just the Outcome: Rewarding Latent Thought Trajectories Improves Reasoning in Looped Language Models2026Amortized Reasoning Tree Search: Decoupling Proposal and Decision in Large Language Models2026PRPO: Aligning Process Reward with Outcome Reward in Policy Optimization2026TraPO: A Semi-Supervised Reinforcement Learning Framework for Boosting LLM Reasoning2025Can LLMs Guide Their Own Exploration? Gradient-Guided Reinforcement Learning for LLM Reasoning2025ScRPO: From Errors to Insights2025Masked-and-Reordered Self-Supervision for Reinforcement Learning from Verifiable Rewards2025GRPO-$\lambda$: Credit Assignment improves LLM Reasoning2025h1: Bootstrapping LLMs to Reason over Longer Horizons via Reinforcement Learning2025MATH-Beyond: A Benchmark for RL to Expand Beyond the Base Model2025Reasoning with Sampling: Your Base Model is Smarter Than You Think2025Inpainting-Guided Policy Optimization for Diffusion Large Language Models2025SPEC-RL: Accelerating On-Policy Reinforcement Learning with Speculative Rollouts2025From Static to Dynamic: Adaptive Monte Carlo Search for Mathematical Process Supervision2025Enhancing Math Reasoning in Small-sized LLMs via Preview Difficulty-Aware Intervention2025wd1: Weighted Policy Optimization for Reasoning in Diffusion Language Models2025Confidence Is All You Need: Few-Shot RL Fine-Tuning of Language Models2025Spurious Rewards: Rethinking Training Signals in RLVR2025$\texttt{SPECS}$: Faster Test-Time Scaling through Speculative Drafts2025Not All Thoughts are Generated Equal: Efficient LLM Reasoning via Multi-Turn Reinforcement Learning2025Maximizing Confidence Alone Improves Reasoning2025Optimal Policy Minimum Bayesian Risk2025SPC: Evolving Self-Play Critic via Adversarial Games for LLM Reasoning2025Towards Hierarchical Multi-Step Reward Models for Enhanced Reasoning in Large Language Models2025Controlling Large Language Model with Latent Actions2025On the Emergence of Thinking in LLMs I: Searching for the Right Intuition2025Exploring the Limit of Outcome Reward for Learning Mathematical Reasoning2025