AI Paper Atlas

Research map / Reinforcement learning / Visual and representation RL

Persistent Robot World Models: Stabilizing Multi-Step Rollouts via Reinforcement Learning

Jai Bardhan, Patrik Drozdík, Josef Sivic, Vladimir Petrik

ECCV 2026 · Reinforcement learning › Visual and representation RL · approach: Vision-based robot RL · 13 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 RL post-training on own rollouts to Ctrl-WorldPersistWorld · DROID val, external camerasPSNR: 23.02 -> 24.42Improved
add RL post-training on own rollouts to Ctrl-WorldPersistWorld · DROID val, external camerasLPIPS: 0.081 -> 0.070Improved
add RL post-training on own rollouts to Ctrl-WorldPersistWorld · DROID val, wrist cameraSSIM: 0.62 -> 0.67Improved
add RL post-training on own rollouts to Ctrl-WorldPersistWorld · DROID val, wrist cameraPSNR: 17.80 -> 19.39Improved
add RL post-training to Ctrl-World, object-only maskedPersistWorld · DROID val, external camera object-onlyLPIPS: 0.025 -> 0.021Improved
add RL post-training to Ctrl-World, robot-only maskedPersistWorld · DROID val, robot-onlyPSNR: 17.62 -> 19.25Improved
replace Self-Forcing style regression with the RL contrastive objective (ablation)PersistWorld · DROID val, 10-step rollout, externalPSNR: 25.35 -> 25.52Improved
replace Self-Forcing style regression with the RL contrastive objective (ablation)PersistWorld · DROID val, 10-step rollout, wristPSNR: 20.17 -> 20.40Improved
replace Self-Forcing style regression with the RL contrastive objective (ablation)PersistWorld · DROID val, 10-step rollout, wristLPIPS: 0.338 -> 0.338No clear effect
lower learning rate from 3e-4 to 1e-4 (ablation)PersistWorld · DROID val, 10-step rollout, externalPSNR: 25.02 -> 25.52Improved
use PersistWorld instead of Ctrl-World for policy evaluationPersistWorld · Policy evaluation (3 tasks, 3 policies)MMRV: 0.053 -> 0.006Improved
use PersistWorld-generated data to improve pi0PersistWorld · Real-robot block pickingtask progression: 0.558 -> 0.667Improved

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