AI Paper Atlas

Research map / Reinforcement learning / Visual and representation RL

Learning Diffusion Policy from Primitive Skills for Robot Manipulation

Zhihao Gu, Ming Yang, Difan Zou, Dong Xu

AAAI 2026 · Reinforcement learning › Visual and representation RL · approach: Dexterous manipulation transfer · 4 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add cross-attention for vision-language injection (ablation)DP baseline · LIBERO-Longsuccess rate (%): 50.5 -> 81.0Improved
add cross-attention for vision-language injection (ablation)DP baseline · CALVIN ABC->Daverage length: 1.13 -> 3.59Improved
add AdaLN prior injection of time and proprioception (ablation)DP + cross-attention · LIBERO-Longsuccess rate (%): 81.0 -> 86.0Improved
add AdaLN prior injection of time and proprioception (ablation)DP + cross-attention · CALVIN ABC->Daverage length: 3.59 -> 4.01Improved
add primitive skill abstraction (skill-dependent FFN) (ablation)DP + prior injection · LIBERO-Longsuccess rate (%): 86.0 -> 91.5Improved
add primitive skill abstraction (skill-dependent FFN) (ablation)DP + prior injection · CALVIN ABC->Daverage length: 4.01 -> 4.32Improved
add compositional prompt ensemble (ablation)SDP without CPE · LIBERO-Longsuccess rate (%): 91.5 -> 93.8Improved
add compositional prompt ensemble (ablation)SDP without CPE · CALVIN ABC->Daverage length: 4.32 -> 4.49Improved
parameterize FFN by skill instead of FiLM conditioning (ablation)SDP · LIBERO-Longsuccess rate (%): 92.5 -> 93.8Improved
parameterize FFN by skill instead of element-wise addition (ablation)SDP · LIBERO-Longsuccess rate (%): 90.9 -> 93.8Improved
parameterize FFN by skill instead of concatenation (ablation)SDP · CALVIN ABC->Daverage length: 4.24 -> 4.49Improved
use SDP instead of MoDE (larger model)diffusion policy · CALVIN ABC->Daverage length: 3.92 -> 4.49Mixed

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy