Stabilizing Long-term Multi-turn Reinforcement Learning with Gated Rewards

Documenti analoghi

• Dynamic Long Context Reasoning over Compressed Memory via End-to-End Reinforcement Learning
  di: Chen, Zhuoen, et al.
  Pubblicazione: (2026)
• Improving Value-based Process Verifier via Structural Prior Injection
  di: Sun, Zetian, et al.
  Pubblicazione: (2025)
• LycheeCluster: Efficient Long-Context Inference with Structure-Aware Chunking and Hierarchical KV Indexing
  di: Li, Dongfang, et al.
  Pubblicazione: (2026)
• LycheeDecode: Accelerating Long-Context LLM Inference via Hybrid-Head Sparse Decoding
  di: Lin, Gang, et al.
  Pubblicazione: (2026)
• Improving Value-based Process Verifier via Low-Cost Variance Reduction
  di: Sun, Zetian, et al.
  Pubblicazione: (2025)