AI Paper Atlas

Research map / Video / Video generation

MemRoPE: Training-Free Infinite Video Generation via Evolving Memory Tokens

Youngrae Kim, Qixin Hu, C.-C. Jay Kuo, Peter Beerel

ECCV 2026 · Video › Video generation · approach: Autoregressive long video generation · 26 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 MemRoPE dual-EMA memory tokens and Online RoPE Indexing (training-free)Self-Forcing · VBench-Long 240 saverage score: 83.04 -> 84.89Improved
add MemRoPELongLive · VBench-Long 1 houraverage score: 87.01 -> 87.99 (vs infty-RoPE)Improved
add Deep Forcing participative compressionSelf-Forcing · VBench-Long 240 saverage score: 83.04 -> 81.66Hurt
add infty-RoPESelf-Forcing · VBench-Long 120 saverage score: 84.25 -> 83.64Hurt
add Deep ForcingLongLive · VBench-Long 240 saverage score: 85.33 -> 85.48No clear effect
add any memory stream (long, short or both) vs no memory (ablation)MemRoPE on LongLive · VBench-Long 60 saverage score: 85.48 -> 85.97Improved
add long-term memory to short-term memory (ablation)MemRoPE on LongLive · VBench-Long 60 saverage score: 85.95 -> 85.97Improved
store keys without RoPE before EMA aggregation instead of aggregating RoPE-rotated keys (ablation)MemRoPE · VBench-Long 60 s Self-Forcingaverage score: 85.28 -> 85.41Improved
store keys without RoPE before EMA aggregation instead of rotated keys (ablation)MemRoPE · VBench-Long 60 s LongLiveaverage score: 86.08 -> 86.24Improved
vary EMA decay rates alpha_L and alpha_S across 12 combinations (ablation)MemRoPE · VBench-Longaverage score: <0.7 variationNo clear effect
use compact three-tier cache (C=12) in place of sliding window (C=21) for Self-ForcingMemRoPE · inference speedvideo FPS: 3.631 -> 5.115Improved
add MemRoPE to LongLive at equal cache sizeLongLive · inference speedvideo FPS: 4.409 -> 4.376No clear effect

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