Tag: Tool Integration

All the papers with the tag "Tool Integration".

• Agent RL Scaling Law: Agent RL with Spontaneous Code Execution for Mathematical Problem Solving

  grok-3-latest

  Score: 0.69

  Published:2025年5月12日 at  at 17:23

  #LLM, #Reinforcement Learning, #Tool Integration, #Mathematical Reasoning, #Scaling Law

  本文通过ZeroTIR框架，揭示了Agent RL Scaling Law，验证了基础LLM可通过强化学习自发学习代码执行工具，显著提升数学推理能力。