SPO: Multi-Dimensional Preference Sequential Alignment With Implicit Reward Modeling

Documenti analoghi

• Uncertainty-aware Reward Model: Teaching Reward Models to Know What is Unknown
  di: Lou, Xingzhou, et al.
  Pubblicazione: (2024)
• Safe Reinforcement Learning with Free-form Natural Language Constraints and Pre-Trained Language Models
  di: Lou, Xingzhou, et al.
  Pubblicazione: (2024)
• Reward-Robust RLHF in LLMs
  di: Yan, Yuzi, et al.
  Pubblicazione: (2024)
• Calibration-Aware Policy Optimization for Reasoning LLMs
  di: Wang, Ziqi, et al.
  Pubblicazione: (2026)
• SPO: Sequential Monte Carlo Policy Optimisation
  di: Macfarlane, Matthew V, et al.
  Pubblicazione: (2024)