AI Paper Atlas

Research map / Video / Video generation

AR-CoPO: Align Autoregressive Video Generation with Contrastive Policy Optimization

Dailan He, Guanlin Feng, Xingtong Ge, Yi ZHANG, Bingqi Ma, Guanglu Song, Yu Liu, Hongsheng LI

ECCV 2026 · Video › Video generation · approach: Autoregressive long video generation · 3 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 semi-on-policy alignmentSelf-Forcing · VBenchTotal: 82.15 -> 82.45Improved
add merged on-policy and semi-on-policy LoRA (scale 0.8)Self-Forcing · VideoAlignOverall: 7.76 -> 8.22Improved
add merged on-policy and semi-on-policy LoRA (scale 0.8)Self-Forcing · VBenchTotal: 82.15 -> 82.17No clear effect
add on-policy AR-CoPO aloneSelf-Forcing · VideoAlignOverall: 7.76 -> 8.51Improved
add on-policy AR-CoPO aloneSelf-Forcing · VBenchTotal: 82.15 -> 81.99Hurt
on-policy training with TA-only reward (ablation)AR-CoPO · VideoAlignMQ: 1.68 -> 0.25Hurt
on-policy training with TA-only reward (ablation)AR-CoPO · VBenchTotal: 82.15 -> 79.26Hurt
on-policy training with TA-only reward (ablation)AR-CoPO · VideoAlignTA: 2.28 -> 2.63Improved
add ratio clipping (semi-on-policy vs fully off-policy) (ablation)AR-CoPO · VBenchTotal: 67.99 -> 82.45Improved
add ratio clipping (semi-on-policy vs fully off-policy) (ablation)AR-CoPO · VideoAlignOverall: 4.23 -> 7.61Improved
increase on-policy LoRA merge scale from 0 to 1.0 (ablation)AR-CoPO merged · VideoAlignOverall: 7.61 -> 8.33Improved
increase on-policy LoRA merge scale from 0 to 1.0 (ablation)AR-CoPO merged · VBenchTotal: 82.45 -> 81.99Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 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