UNA: A Unified Supervised Framework for Efficient LLM Alignment Across Feedback Types

Documenti analoghi

• PAFT: A Parallel Training Paradigm for Effective LLM Fine-Tuning
  di: Pentyala, Shiva Kumar, et al.
  Pubblicazione: (2024)
• Reinforcement Learning for LLM Post-Training: A Survey
  di: Wang, Zhichao, et al.
  Pubblicazione: (2024)
• Diversity Enhances an LLM's Performance in RAG and Long-context Task
  di: Wang, Zhichao, et al.
  Pubblicazione: (2025)
• UFT: Unifying Fine-Tuning of SFT and RLHF/DPO/UNA through a Generalized Implicit Reward Function
  di: Wang, Zhichao, et al.
  Pubblicazione: (2024)
• Rate, Explain and Cite (REC): Enhanced Explanation and Attribution in Automatic Evaluation by Large Language Models
  di: Hsu, Aliyah R., et al.
  Pubblicazione: (2024)