$π$-StepNFT: Wider Space Needs Finer Steps in Online RL for Flow-based VLAs

Documenti analoghi

• SpatialViz-Bench: A Cognitively-Grounded Benchmark for Diagnosing Spatial Visualization in MLLMs
  di: Wang, Siting, et al.
  Pubblicazione: (2025)
• Shallow-π: Knowledge Distillation for Flow-based VLAs
  di: Jeon, Boseong, et al.
  Pubblicazione: (2026)
• SnapFlow: One-Step Action Generation for Flow-Matching VLAs via Progressive Self-Distillation
  di: Luan, Wuyang, et al.
  Pubblicazione: (2026)
• Finer Parameter Steps for Low-Rank PEFT: A Controlled Study with CP Tensor Adapters
  di: Wang, Xinjue, et al.
  Pubblicazione: (2026)
• $π_\texttt{RL}$: Online RL Fine-tuning for Flow-based Vision-Language-Action Models
  di: Chen, Kang, et al.
  Pubblicazione: (2025)