AI Paper Atlas

Research map / Video / Video generation

Cosmos Policy: Fine-Tuning Video Models for Visuomotor Control and Planning

Moo Jin Kim, Yihuai Gao, Tsung-Yi Lin, Yen-Chen Lin, Yunhao Ge, Grace Lam, Percy Liang, Shuran Song, Ming-Yu Liu, Chelsea Finn, Jinwei Gu

ICLR 2026 · Video › Video generation · approach: World-model video generation · 353 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
initialize from pretrained video model instead of random weights (ablation)Cosmos Policy · LIBERO (average)success rate (%): 94.6 -> 98.5Improved
initialize from pretrained video model instead of random weights (ablation)Cosmos Policy · ALOHA fold shirtscore: 80.8 -> 18.7 points higher full modelImproved
add auxiliary future-state and value prediction targets (ablation)Cosmos Policy · LIBERO (average)success rate (%): 97.0 -> 98.5Improved
add value function training samples (ablation)Cosmos Policy · RoboCasa (average)success rate (%): 66.6 -> 67.1Improved
add world model training samples (ablation)Cosmos Policy · RoboCasa (average)success rate (%): 64.0 -> 66.6Improved
add model-based planning with V(s') using refined world model (ablation)Cosmos Policy · ALOHA two hardest tasksaverage score: +12.5 pointsImproved
plan with state value V(s') via world model instead of model-free Q(s,a) (ablation)Cosmos Policy planning · ALOHA two hardest tasksscoreImproved
fine-tune pretrained video model into policy instead of fine-tuned VLAsLIBERO · LIBERO (average)success rate (%): 97.4 -> 98.5 (vs CogVLA)Improved
use Cosmos Policy instead of pi_0.5 on out-of-distribution ALOHA testsALOHA · Real-world ALOHA (OOD only)average score: 92.5 -> 89.3Hurt

Later papers that compared with this method

3 later accepted papers report Cosmos-Policy as a baseline in their results tables; 3 of them report a better number on most of the results they share with it.

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