D5RL: Diverse Datasets for Data-Driven Deep Reinforcement Learning

Documenti analoghi

• Efficient Imitation Learning with Conservative World Models
  di: Kolev, Victor, et al.
  Pubblicazione: (2024)
• Cal-QL: Calibrated Offline RL Pre-Training for Efficient Online Fine-Tuning
  di: Nakamoto, Mitsuhiko, et al.
  Pubblicazione: (2023)
• MOTO: Offline Pre-training to Online Fine-tuning for Model-based Robot Learning
  di: Rafailov, Rafael, et al.
  Pubblicazione: (2024)
• Preference Fine-Tuning of LLMs Should Leverage Suboptimal, On-Policy Data
  di: Tajwar, Fahim, et al.
  Pubblicazione: (2024)
• Training Diffusion Models with Reinforcement Learning
  di: Black, Kevin, et al.
  Pubblicazione: (2023)