AI Paper Atlas

Research map / Video / Video generation

LongLive: Real-time Interactive Long Video Generation

Shuai Yang, Wei Huang, Ruihang Chu, Yicheng Xiao, Yuyang Zhao, Xianbang Wang, Muyang Li, Enze Xie, Ying-Cong Chen, Yao Lu, Song Han, Yukang Chen

ICLR 2026 · Video › Video generation · approach: Autoregressive long video generation

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 KV recache at prompt switches instead of retaining the full KV cache (ablation)LongLive · Prompt switch at 5s (10s video)CLIP score: 25.92 -> 27.87Improved
add KV recache instead of clearing the cache at prompt switches (ablation)LongLive · Prompt switch at 5s (10s video)subject consistency: 89.59 -> 94.04Improved
KV recache instead of clearing cache (ablation)LongLive · Prompt switch at 5s (10s video)CLIP score: 28.95 -> 27.87Hurt
add short-window attention with frame sinkLongLive inference · single H100, long videocompute time and peak memory: -28% compute time, -17% peak memoryImproved
add frame sink to short-window attention (9 local + 3 sink) (ablation)LongLive · 10s generation consistencybackground and subject consistencyImproved
increase attention window from 3 to 27 latent frames (ablation)LongLive · 10s generation consistencyconsistencyImproved
train long (streaming long tuning) instead of train-short-test-longLongLive vs Self-Forcing · VBench-Long 30stotal score: 81.59 -> 83.52Improved
increase LoRA rank from 32 to 256 (ablation)streaming long tuning · VBench-Long 30stotal score: 81.08 -> 83.12Improved
increase LoRA rank from 256 to 512 (ablation)streaming long tuning · VBench-Long 30stotal score: 83.12 -> 83.04No clear effect
full fine-tuning instead of LoRA rank 256 (ablation)streaming long tuning · VBench-Long 30stotal score: 83.12 -> 83.52Improved
INT8 post-training quantizationLongLive · VBenchtotal score: 84.87 -> 84.31; 12.6 -> 16.4 FPSMixed
recaching cost at prompt switch during trainingLongLive · 10s video with one switchextra time cost: about 6% extra timeNo clear effect

Later papers that compared with this method

34 later accepted papers report LongLive as a baseline in their results tables; 33 of them report a better number on most of the results they share with it.

26 more not listed.

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