AI Paper Atlas

Research map / Reinforcement learning / Visual and representation RL

VPA-WM: Vision-Priors-Aligned World Models for Robust Visual Reinforcement Learning

Xu Zhang, Sicong Liu, Liwei Guo, Chenjuan Guo, Bin Yang, Yang Shu

ECCV 2026 · Reinforcement learning › Visual and representation RL · approach: Robust visual encoders for RL · 0 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 control-oriented alignment fine-tuning of PVM (ablation)DA-WM · Distracted DMC Walker Runepisode return: 424 -> 572Improved
add instant action prediction (ablation)DA-WM · Distracted DMC Walker Runepisode return: 219 -> 572Improved
add short-horizon return estimation (ablation)DA-WM · Distracted DMC Cheetah Runepisode return: 402 -> 804Improved
add spatially grounded guidance (ablation)DA-WM · Distracted DMC Cheetah Runepisode return: 154 -> 804Improved
add experience query integration (ablation)DA-WM · Distracted DMC Walker Runepisode return: 478 -> 572Improved
overlay DINO masks on input instead of STCI (ablation)DA-WM · Distracted DMC Walker Runepisode return: 572 -> 146Hurt
replace world model encoder with DINO (ablation)DA-WM · Distracted DMC Cheetah Runepisode return: 804 -> 349Hurt
use DA-WM instead of Dreamer under distractionsDreamer · Distracted DMC Cheetah Runepisode return: 104 -> 804Improved
use DA-WM instead of Dreamer on clean Standard DMCDA-WM · Standard DMC Cheetah Runepisode return: 892 -> 789No clear effect
swap DINOv2 for CLIP as PVM (CA-WM)VPA-WM · Distracted DMC Cheetah Runepisode return: 804 -> 699Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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