Trajectory Balance with Asynchrony: Decoupling Exploration and Learning for Fast, Scalable LLM Post-Training

Documenti analoghi

• Recursive Self-Aggregation Unlocks Deep Thinking in Large Language Models
  di: Venkatraman, Siddarth, et al.
  Pubblicazione: (2025)
• A Comedy of Estimators: On KL Regularization in RL Training of LLMs
  di: Shah, Vedant, et al.
  Pubblicazione: (2025)
• Adversarial Robustness Limits via Scaling-Law and Human-Alignment Studies
  di: Bartoldson, Brian R., et al.
  Pubblicazione: (2024)
• Speculative Diffusion Decoding: Accelerating Language Generation through Diffusion
  di: Christopher, Jacob K, et al.
  Pubblicazione: (2024)
• Double Visual Defense: Adversarial Pre-training and Instruction Tuning for Improving Vision-Language Model Robustness
  di: Wang, Zeyu, et al.
  Pubblicazione: (2025)