AIME-24/25
Emerging4papers using it
2025first seen
The 'AIME-24/25' dataset/benchmark contains a collection of tasks designed to evaluate the performance and robustness of reinforcement learning algorithms, particularly in the context of agentic problem-solving with Large Language Models.
Papers using AIME-24/25 (4)
- PopuLoRA: Co-Evolving LLM Populations for Reasoning Self-PlayCLEANER: Self-Purified Trajectories Boost Agentic Reinforcement LearningTraPO: A Semi-Supervised Reinforcement Learning Framework for Boosting LLM ReasoningSample More to Think Less: Group Filtered Policy Optimization for Concise Reasoning