AI Paper Atlas

Research map / Reinforcement learning / Visual and representation RL

Real-Time Robot Execution with Masked Action Chunking

Haoxuan Wang, Gengyu Zhang, Yan Yan, Yuzhang Shang, Ramana Rao Kompella, Gaowen Liu

ICLR 2026 · Reinforcement learning › Visual and representation RL · approach: Dexterous manipulation transfer · 18 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 LoRA alone without masked training (ablation)pi0/flow policy · Kinetixsolve rateNo clear effect
progressively add masked chunking and prefix-preserved sampling components (ablation)REMAC · Kinetixsolve rateImproved
use self-conditioned curriculum schedule instead of pure ground-truth or pure self-conditioned inputs (ablation)REMAC · Kinetixsolve rate and stabilityImproved
add BID at test time (ablation)REMAC · Kinetix d=4solve rate: 0.779 -> 0.781Improved
add RTC at test time (ablation)REMAC · Kinetix d=4solve rate: 0.779 -> 0.791Improved
increase q_min / q_max (ablation)REMAC · Kinetixsolve rateNo clear effect
add learned mask embedding (ablation)REMAC · Kinetixsolve rateNo clear effect
reduce training data from 200 to 30 trajectories (ablation)REMAC · real robot Grasp-Hardcompletion progress: 0.820 -> 0.810No clear effect
use REMAC vs RTCREMAC · Kinetix d=4solve rate: 0.588 -> 0.760Improved
add REMAC to ACT vs naive asyncACT · Transfer Cube d=4success rate: 0.40 -> 0.74Improved
add REMAC to ACTACT · Insert Box d=10success rate: 0.14 -> 0.16Mixed
corrupt delay input with noisy and spiky valuesREMAC · Kinetix d=3solve rate: 0.817 -> 0.718 (d=3 spiky)Hurt

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