2 papers
cs.CV2026
TailProp: content-adaptive light- and heavy-tailed propagation for vision
Jiahao Kong, Zihan Li
Science-inspired vision models show that explicit propagation dynamics can provide structured and interpretable alternatives to conventional token mixing. Existing formulations, ho…
cs.CV2026
LayerRecall: A State-Conditioned Memory Router for Long-Horizon Consistency in Video Generation
Yixuan Ding, Jiahao Kong, Wei Huang +2
Autoregressive video diffusion enables scalable long-video generation by producing chunks from a bounded recent context. While recency-based caching preserves local continuity, it…