Trendoday

Disen-Forcing: Disentangling Semantic Anchoring from Motion for Autoregressive Video Diffusion

Autoregressive video diffusion models (AR-VDMs) enable real-time, long-horizon video generation but degrade rapidly once the inference horizon outruns the training horizon, due to compounding exposure

  • Upvotes5
–Rank now
#27Peak rank
2Days on list
15Score · 7d
Nov 16, 2026First seen
Rank · 7 days

About this paper

Disen-Forcing is a distillation-time method for autoregressive video diffusion models that disentangles attention-head specialization to improve long-range consistency and motion dynamics in long-horizon video generation.

Autoregressive video diffusion models (AR-VDMs) enable real-time, long-horizon video generation but degrade rapidly once the inference horizon outruns the training horizon, due to compounding exposure bias. To stabilize long rollouts, recent methods heuristically adopt frame sink, keeping initial frames as a persistent model context, inspired by attention sinks in LLMs. We revisit this design and find that the mechanism by which initial frames stabilize long-range consistency in AR-VDMs is not an LLM-style attention sink; instead, long-range consistency is anchored through a sparse subset of attention heads, while the majority of heads predominantly capture local temporal dynamics. This…

Autoregressive video diffusion modelsAR-VDMslong-horizon video generationinference horizontraining horizonexposure biaslong rolloutsframe sinkattention sinkslong-range consistency
Upvotes
5
Authors
8
Published
Nov 2, 2026

Summary and abstract excerpt via Hugging Face Papers ↗ · arXiv.

Trending history

Disen-Forcing: Disentangling Semantic Anchoring from Motion for Autoregressive Video Diffusion was first seen on Hugging Face · Daily papers on Nov 16, 2026. Since then it has appeared in 6 checks on 2 different days. Its best position was #27 of 30, reached on Nov 17, 09:28 UTC. It dropped off the list after Nov 17, 09:28 UTC.

Rank history

Position on Hugging Face · Daily papers at every check (lower is better).

#1#11#20#30 07:2813:5820:2802:5809:28 Nov 16, 07:28 UTC · #30Nov 16, 15:28 UTC · #30Nov 16, 23:28 UTC · #30Nov 17, 01:28 UTC · #29Nov 17, 03:28 UTC · #30Nov 17, 09:28 UTC · #27

Last 14 days

Best rank per day (UTC).

Nov 9today

Metric history

Platform metric as reported by Hugging Face.

0369 07:2813:5820:2802:5809:28 Nov 16, 07:28 UTC · 5Nov 16, 15:28 UTC · 9Nov 16, 23:28 UTC · 9Nov 17, 01:28 UTC · 9Nov 17, 03:28 UTC · 9Nov 17, 09:28 UTC · 9

Questions

Is Disen-Forcing: Disentangling Semantic Anchoring from Motion for Autoregressive Video Diffusion trending on HF Daily Papers right now?

Not at the moment. It was last on the list on Nov 17, 09:28 UTC, after 2 days of trending.

What is the highest rank Disen-Forcing: Disentangling Semantic Anchoring from Motion for Autoregressive Video Diffusion reached?

#27 on Hugging Face · Daily papers, on Nov 17, 09:28 UTC. Trendoday checks this list every 120 minutes.

How long has Disen-Forcing: Disentangling Semantic Anchoring from Motion for Autoregressive Video Diffusion been trending?

It was first seen on Nov 16, 2026 and has been on the list on 2 days (6 checks) since.