AIME 2025

Homepage and repository Homepage: https://matharena.ai/ Repository: https://github.com/eth-sri/matharena Dataset Summary This dataset contains the questions from AIME 2025 used for the MathArena Leaderboard Data Fields The dataset contains the following fields: problem_idx (int64): Problem index within the correspondin

Papers using AIME 2025 (23)

Every Step Evolves: Scaling Reinforcement Learning for Trillion-Scale Thinking Model2025 · 20 citesUloRL:An Ultra-Long Output Reinforcement Learning Approach for Advancing Large Language Models' Reasoning Abilities2025 · 9 citesGEPA: Reflective Prompt Evolution Can Outperform Reinforcement Learning2025 · 1 citesExpanding Reasoning Potential in Foundation Model by Learning Diverse Chains of Thought Patterns2025 · 2 citesLearning to Reason by Analogy via Retrieval-Augmented Reinforcement Fine-Tuning2026Rollout Pass-Rate Control: Steering Binary-Reward RL Toward Its Most Informative Regime2026Maximizing Rollout Informativeness under a Fixed Budget: A Submodular View of Tree Search for Tool-Use Agentic Reinforcement Learning2026expo: Exploration-prioritized policy optimization via adaptive kl regulation and gaussian curriculum sampling2026fg-expo: Frontier-guided exploration-prioritized policy optimization via adaptive kl and gaussian curriculum2026Apriel-1.5-OpenReasoner: RL Post-Training for General-Purpose and Efficient Reasoning2026Good SFT Optimizes for SFT, Better SFT Prepares for Reinforcement Learning2026ForesightKV: Optimizing KV Cache Eviction for Reasoning Models by Learning Long-Term Contribution2026Beyond Variance: Prompt-Efficient RLVR via Rare-Event Amplification and Bidirectional Pairing2026Training Large Reasoning Models Efficiently via Progressive Thought Encoding2026PCL-Reasoner-V1.5: Advancing Math Reasoning with Offline Reinforcement Learning2026MarsRL: Advancing Multi-Agent Reasoning System via Reinforcement Learning with Agentic Pipeline Parallelism2025Plan Then Action:High-Level Planning Guidance Reinforcement Learning for LLM Reasoning2025Unlocking Reasoning Capabilities in LLMs via Reinforcement Learning Exploration2025BAPO: Stabilizing Off-Policy Reinforcement Learning for LLMs via Balanced Policy Optimization with Adaptive Clipping2025EAPO: Enhancing Policy Optimization with On-Demand Expert Assistance2025Kimi K2: Open Agentic Intelligence2025Advancing Multimodal Reasoning: From Optimized Cold Start to Staged Reinforcement Learning2025AceReason-Nemotron: Advancing Math and Code Reasoning through Reinforcement Learning2025