AI Paper Atlas

Research map / Video / Video generation

Deep Forcing: Training-Free Long Video Generation with Deep Sink and Participative Compression

Jung Yi, Wooseok Jang, Paul Hyunbin Cho, Jisu Nam, Heeji Yoon, Seungryong Kim

ICML 2026 · Video › Video generation · approach: Autoregressive long video generation · 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 Deep Sink to Self Forcing (ablation)Self Forcing · VBench-Long 30sDynamic Degree: 36.62 -> 48.58Improved
add Participative Compression on top of Deep Sink (ablation)SF + Deep Sink · VBench-Long 30sDynamic Degree: 48.58 -> 57.56Improved
add Participative Compression on top of Deep Sink (ablation)SF + Deep Sink · VBench-Long 30sImage Quality: 68.54 -> 69.31Improved
add Deep Sink to Self Forcing (ablation)Self Forcing · VBench-Long 30sImage Quality: 68.58 -> 68.54No clear effect
add Deep Sink and Participative Compression to Self Forcing (ablation)Self Forcing · VBench-Long 30sOverall Consistency: 20.50 -> 20.54Improved
select Top-C with both past and denoising queries instead of only denoising (ablation)Participative Compression · VBench-Long 30sImage Quality: 68.24 -> 69.31Improved
select Top-C with both past and denoising queries instead of only past (ablation)Participative Compression · VBench-Long 30sImage Quality: 68.54 -> 69.31Improved
select Top-C with both queries instead of only denoising (ablation)Participative Compression · VBench-Long 30sMotion Smoothness: 97.86 -> 98.27Improved
apply Deep Forcing training-free to Self ForcingSelf Forcing · VBench-Long 60sDynamic Degree: 31.98 -> 57.19Improved
apply Deep Forcing training-free to Self ForcingSelf Forcing · VBench-Long 60sImaging Quality: 66.33 -> 69.27Improved
apply Deep Forcing to Self ForcingSelf Forcing · VLM visual stabilityvisual stability: 43.94 -> 75.44Improved
add Deep Forcing compression to Self ForcingSelf Forcing · 60s generation, H100throughput FPS: 15.78 -> 15.75No clear effect

Later papers that compared with this method

7 later accepted papers report Deep Forcing as a baseline in their results tables; 7 of them report a better number on most of the results they share with it.

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 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