What training and inference remedies do the primary papers use against error accumulation in long autoregressive video, how does each work in steps, and what does each cost?
Seven remedy families appear: (1) noise the past latent at sampling (Diffusion Forcing); (2) distil a causal student from a bidirectional teacher with DMD (CausVid); (3) train on self-generated rollouts with KV cache (Self Forcing); (4) roll out beyond the teacher horizon and distil on re-noised windows (Self-Forcing++); (5) train-long-test-long with a frame sink (LongLive); (6) rolling-window joint denoising with an attention sink (Rolling Forcing); (7) error recycling (Stable Video Infinity). The papers disagree on whether an attention sink works alone (LongLive says only after long tuning; Rolling Forcing ablates it as helpful in its own recipe).
Seven remedy families appear: (1) noise the past latent at sampling (Diffusion Forcing); (2) distil a causal student from a bidirectional teacher with DMD (CausVid); (3) train on self-generated rollouts with KV cache (Self Forcing); (4) roll out beyond the teacher horizon and distil on re-noised windows (Self-Forcing++); (5) train-long-test-long with a frame sink (LongLive); (6) rolling-window joint denoising with an attention sink (Rolling Forcing); (7) error recycling (Stable Video Infinity). The papers disagree on whether an attention sink works alone (LongLive says only after long tuning; Rolling Forcing ablates it as helpful in its own recipe).
Curator qualifications: All excerpts located; attention-sink efficacy disagreement across LongLive/Rolling Forcing/Self-Forcing++ kept.
Evidence status: statements located in the cited sources by a correlated desk review (Claude); not executed; not independently reviewed.
Structured fields
{
"citations": [
{
"url": "https://arxiv.org/abs/2407.01392v4",
"note": "candidate evidence excerpts"
},
{
"url": "https://arxiv.org/abs/2412.07772v4",
"note": "candidate evidence excerpts"
},
{
"url": "https://arxiv.org/abs/2506.08009v2",
"note": "candidate evidence excerpts"
},
{
"url": "https://arxiv.org/abs/2510.02283v1",
"note": "candidate evidence excerpts"
},
{
"url": "https://arxiv.org/abs/2509.22622v2",
"note": "candidate evidence excerpts"
},
{
"url": "https://arxiv.org/abs/2509.25161v1",
"note": "candidate evidence excerpts"
},
{
"url": "https://arxiv.org/abs/2510.09212v1",
"note": "candidate evidence excerpts"
}
]
}Evidence claims
- [sourced_fact] Seven remedy families appear: (1) noise the past latent at sampling (Diffusion Forcing); (2) distil a causal student from a bidirectional teacher with DMD (CausVid); (3) train on self-generated rollouts with KV cache (Self Forcing); (4) roll out beyond the teacher horizon and distil on re-noised windows (Self-Forcing++); (5) train-long-test-long with a frame sink (LongLive); (6) rolling-window joint denoising with an attention sink (Rolling Forcing); (7) error recycling (Stable Video Infinity). The papers disagree on whether an attention sink works alone (LongLive says only after long tuning; Rolling Forcing ablates it as helpful in its own recipe). · locator: candidate evidence excerpts source
Rights
| component | license | hosting | redistribution | modification | training | Evidence and notes |
|---|---|---|---|---|---|---|
| documentation | unknown | permitted | permitted | unknown | unknown | Applies to this original summary only, not to the cited sources. |