SPIRAL: Self-Play on Zero-Sum Games Incentivizes Reasoning via Multi-Agent Multi-Turn Reinforcement Learning

Documenti analoghi

• TextArena
  di: Guertler, Leon, et al.
  Pubblicazione: (2025)
• Chasing Moving Targets with Online Self-Play Reinforcement Learning for Safer Language Models
  di: Liu, Mickel, et al.
  Pubblicazione: (2025)
• STLM Engineering Report: Dropout
  di: Hillier, Dylan, et al.
  Pubblicazione: (2024)
• Consistently Simulating Human Personas with Multi-Turn Reinforcement Learning
  di: Abdulhai, Marwa, et al.
  Pubblicazione: (2025)
• Super Tiny Language Models
  di: Hillier, Dylan, et al.
  Pubblicazione: (2024)