🛰️ Daily AI Frontier
‹ back to 2026-08-31

LayerRecall: A State-Conditioned Memory Router for Long-Horizon Consistency in Video Generation

Research Multimodal & Generative

Ranking

Overall 82
Content 90
Popularity 65

Observed public metrics from 1 member.

Representative image for LayerRecall: A State-Conditioned Memory Router for Long-Horizon Consistency in Video Generation

Merged summary

TL;DR - LayerRecall is a state-conditioned memory router that helps autoregressive video diffusion models recover relevant distant context by injecting retrieved historical K/V states into selected transformer layers. It improves long-horizon consistency with negligible inference overhead while preserving local motion and scene continuity.

  • Identifies distinct layer-level preferences for current, recent, and distant context in video diffusion transformers.
  • Routes relevant historical states only to memory-sensitive layers, leaving local attention unchanged elsewhere.
  • Uses Cross-Horizon Prediction Matching to train from a privileged long-context model without explicit memory-allocation labels.
  • Achieves the best reported overall results on MemoBench and MovieBench across 100 multi-shot prompts while matching its backbone on VBench-Long.

Sources (1)

LayerRecall: A State-Conditioned Memory Router for Long-Horizon Consistency in Video Generation

arXiv cs.CV Yixuan Ding, Jiahao Kong, Wei Huang, Ruijie Quan, Yi Yang 2026-08-28 arXiv:2608.28460
Public signals Hugging Face upvotes 29
Providers: Hugging Face · Upvotes 29 OpenAlex · N/A Publisher · N/A Semantic Scholar · N/A X · N/A Fetched 2026-09-25 14:25:52.301780 UTC

TL;DR - LayerRecall is a state-conditioned memory router that helps autoregressive video diffusion models recover relevant distant context by injecting retrieved historical K/V states into selected transformer layers. It improves long-horizon consistency with negligible inference overhead while preserving local motion and scene continuity.

  • Identifies distinct layer-level preferences for current, recent, and distant context in video diffusion transformers.
  • Routes relevant historical states only to memory-sensitive layers, leaving local attention unchanged elsewhere.
  • Uses Cross-Horizon Prediction Matching to train from a privileged long-context model without explicit memory-allocation labels.
  • Achieves the best reported overall results on MemoBench and MovieBench across 100 multi-shot prompts while matching its backbone on VBench-Long.
item →