Anchor Forcing: Anchor Memory and Tri-Region RoPE for Interactive Streaming Video Diffusion

Yang Yang; Tianyi Zhang; Wei Huang; Jinwei Chen; Boxi Wu; Xiaofei He; Deng Cai; Bo Li

cs.CVMar 12, 2026

Anchor Forcing: Anchor Memory and Tri-Region RoPE for Interactive Streaming Video Diffusion

Authors: Yang Yang, Tianyi Zhang, Wei Huang, Jinwei Chen, Boxi Wu, Xiaofei He, Deng Cai, Bo Li,

Organizations: Zhejiang University · vivo BlueImage Lab · University of Hong Kong

Abstract

Interactive long video generation requires prompt switching to introduce new subjects or events, while maintaining perceptual fidelity and coherent motion over extended horizons. Recent distilled streaming video diffusion models reuse a rolling KV cache for long-range generation, enabling prompt-switch interaction through re-cache at each switch. However, existing streaming methods still exhibit progressive quality degradation and weakened motion dynamics. We identify two failure modes specific to interactive streaming generation: (i) at each prompt switch, current cache maintenance cannot simultaneously retain KV-based semantic context and recent latent cues, resulting in weak boundary conditioning and reduced perceptual quality; and (ii) during distillation, unbounded time indexing induces a positional distribution shift from the pretrained backbone's bounded RoPE regime, weakening pretrained motion priors and long-horizon motion retention. To address these issues, we propose Anchor Forcing, a cache-centric framework with two designs. First, an anchor-guided re-cache mechanism stores KV states in anchor caches and warm-starts re-cache from these anchors at each prompt switch, reducing post-switch evidence loss and stabilizing perceptual quality. Second, a tri-region RoPE with region-specific reference origins, together with RoPE re-alignment distillation, reconciles unbounded streaming indices with the pretrained RoPE regime to better retain motion priors. Experiments on long videos show that our method improves perceptual quality and motion metrics over prior streaming baselines in interactive settings. Project page: https://github.com/vivoCameraResearch/Anchor-Forcing

Figures & tables

Figure 1: We propose Anchor Forcing , which supports prompt switches that introduce new subjects and actions while preserving context, motion quality, and temporal coherence across clips. In contrast, prior methods degrade over time and often fail to realize newly introduced interactions, as highlighted by the red boxes. Red text denotes the interaction newly specified in each segment.

Figure 2: The Anchor Forcing pipeline. (a) Overview of Anchor Forcing in an interactive setting with two prompt switches. We denote the anchor memory for generating frame

t

\mathcal{M}(t)

, and apply anchor-guided re-cache at

f_{1}

and

f_{2}

to update the local KV cache under the new prompt condition. (b) Prior re-cache [ 34 ] . It rebuilds the local cache solely from historical frame latents, which fails to retain prior KV evidence across prompt switches. (c) Anchor-guided re-cache at

f_{2}

. It augments re-cache with the anchor memory

\mathcal{M}(6)

and refreshes the junction caches

x_{5}

Method	Dynamic Degree $\uparrow$	Quality Score $\uparrow$	CLIP Score $\uparrow$
Method	Dynamic Degree $\uparrow$	Quality Score $\uparrow$	0–10s	10–20s	20–30s	30–40s	40–50s	50–60s
Infinity-RoPE [ 35 ]	$35.17$	$79.98$	$23.87$	$22.62$	$22.16$	$21.82$	$22.20$	$22.20$
MemFlow [ 17 ]	$45.47$	$79.35$	24.87	$24.08$	$23.02$	$22.87$	$22.35$	$21.59$
LongLive [ 34 ]	$26.37$	$78.92$	$24.82$	$24.18$	$23.30$	$22.98$	$23.40$	$22.82$
Anchor Forcing	73.00	82.25	$24.48$	24.35	23.50	23.53	23.61	23.63

Table 1: Interactive long video evaluation results. The dynamic degree and quality scores are reported on the whole 60s sequence. CLIP scores are reported on 10s video segments with the same semantics (

\uparrow

higher is better).

Model	#Params	Resolution	Dynamic $\uparrow$	Evaluation scores $\uparrow$
Model	#Params	Resolution	Dynamic $\uparrow$	Total	Quality	Semantic
Real-data-trained models
Pyramid Flow [ 18 ]	2B	$640{\times}384$	68.06	81.56	83.89	72.24
LTX-Video [ 11 ]	1.9B	$768{\times}512$	75.00	82.33	85.20	70.83
SkyReels-V2 [ 3 ]	1.3B	$960{\times}540$	48.61	81.72	82.93	76.90
Wan2.1 [ 29 ]	1.3B	$832{\times}480$	73.61	84.27	85.24	80.43

Table 2: Comparison with relevant baselines under 5s setting. We compare Anchor Forcing with representative open-source models of similar parameter sizes and resolutions. Evaluation scores are calculated on the standard prompt of VBench.

Figure 3: Qualitative comparison on interactive long-video generation. Compared with baselines, Anchor Forcing achieves stronger prompt compliance, more coherent dynamic motion, and higher long-range visual quality. Red text marks the interactive content newly introduced in each segment.

Model	Total Score $\uparrow$	Quality Score $\uparrow$	Semantic Score $\uparrow$	Dynamic Score $\uparrow$	Imaging Quality $\uparrow$	Temporal Style $\uparrow$
Self-Forcing [ 14 ]	82.21	83.38	77.54	51.85	68.17	23.31
Infinity-RoPE [ 35 ]	83.16	84.17	79.13	57.04	67.94	23.79
DeepForcing [ 36 ]	82.26	83.20	78.50	52.86	67.65	23.51
MemFlow [ 17 ]	82.95	83.86	79.31	60.46	67.68	23.89
LongLive [ 34 ]	82.77	83.31	80.64	40.19	68.96	23.97
Anchor Forcing	83.25	84.40	78.67	79.17	69.37	24.17

Table 3: Comparison with relevant baselines under 30s setting. We compare Anchor Forcing with representative open-source models. Evaluation scores are calculated on the standard prompt of VBench-Long.

Figure 4: Qualitative comparison on 30-second long-video generation. Compared with prior methods, ours method yields higher-quality, more prompt-faithful videos, avoiding the background and content degradations highlighted in the yellow and red boxes.

Figure 5: Ablation on 60-second interactive generation. Tri-region RoPE improves motion dynamics, and anchor-guided re-cache further preserves perceptual quality and prompt compliance across segments. Red boxes mark degradation between adjacent clips, and red text denotes newly introduced interactions.

Method	CLIP Score $\uparrow$	Dynamic Degree $\uparrow$	Quality Score $\uparrow$
Baseline	23.58	26.37	78.92
+ tr-RoPE	23.71	74.83	81.72
+ tr-RoPE + AR	23.85	73.00	82.25

Table 4: Ablation study on key components. The best results for the module are indicated in bold.

Video Generation

Video generation studies models that create and edit moving images from text, images, audio, and structured controls.

See all 364 papers

May 15, 2026

Echo-Forcing: A Scene Memory Framework for Interactive Long Video Generation

Autoregressive video diffusion models enable open-ended generation through local attention and KV caching. However, existing training-free long-video optimization methods mainly focus on stable extension under a single prompt, making them difficult to handle interactive scenarios involving prompt switching, old scene forgetting, and historical scene recall. We identify the core bottleneck as the functional entanglement of historical KV states: stable anchors and recent dynamics are handled by the same cache policy, leading to outdated background contamination, delayed response to new prompts, and loss of long-range memory. To address this issue, we propose Echo-Forcing, a training-free scene memory framework specifically designed for interactive long video generation with three core mechanisms: (1) Hierarchical Temporal Memory, which decouples stable anchors, compressed history, and recent windows under relative RoPE; (2) Scene Recall Frames, which compresses historical scenes into spatially structured KV representations to support long-term recall; and (3) Difference-aware Memory Decay, which adaptively forgets conflicting tokens according to the discrepancy between old and new scenes. Based on these designs, Echo-Forcing uniformly supports smooth transitions, hard cuts, and long-range scene recall under a bounded cache budget. Extensive evaluations on VBench-Long further demonstrate that Echo-Forcing achieves the best overall performance in both long-video generation and interactive video generation settings. Our code is released in https://github.com/mingqiangWu/Echo-Forcing

Mingqiang Wu, Weilun Feng, Zhefeng Zhang +8

State Key Laboratory of AI Safety, Institute of Computing Technology, Chinese Academy of Sciences, Beijing, China · University of Chinese Academy of Sciences, Beijing, China · China University of Mining & Technology, Beijing +3

Video Diffusion Models Long-Horizon Video Generation Memory-Augmented Video Generation Video Generation Autoregressive Video Generation

Ask PDF

Mar 12, 2026

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

Autoregressive diffusion enables real-time frame streaming, yet existing sliding-window caches discard past context, causing fidelity degradation, identity drift, and motion stagnation over long horizons. Current approaches preserve a fixed set of early tokens as attention sinks, but this static anchor cannot reflect the evolving content of a growing video. We introduce MemRoPE, a training-free framework with two co-designed components. Memory Tokens continuously compress all past keys into dual long-term and short-term streams via exponential moving averages, maintaining both global identity and recent dynamics within a fixed-size cache. Online RoPE Indexing caches unrotated keys and applies positional embeddings dynamically at attention time, ensuring the aggregation is free of conflicting positional phases. These two mechanisms are mutually enabling: positional decoupling makes temporal aggregation well-defined, while aggregation makes fixed-size caching viable for unbounded generation. Extensive experiments validate that MemRoPE outperforms existing methods in temporal coherence, visual fidelity, and subject consistency across minute- to hour-scale generation.

Youngrae Kim, Qixin Hu, C. -C. Jay Kuo +1

University of Southern California

Long-Horizon Video Generation Memory-Augmented Video Generation Diffusion Model Caching Temporal Consistency in Video Generation Autoregressive Diffusion

Ask PDF

Oct 8, 2026

Memory Forcing: Attendable Mid-Horizon History for Streaming Video Generation

Autoregressive video diffusion enables causal video streaming without a bidirectional pass over the full clip, but existing few-step systems usually retain only the opening and most recent frames in a fixed-size KV cache. Once an event leaves this window, later frames can no longer attend to it, a failure we term mid-horizon forgetting. We present Memory Forcing, a few-step streaming method that preserves this missing history without increasing the cache size. Its Archive & Working Banks partition the cache into sink, archive, and working regions, retaining diverse intermediate events alongside recent motion under fixed memory. Because absolute temporal indices drift outside the training range, Bank-aware RoPE reassigns indices at attention time so each bank remains distinguishable. At 1.3B, Memory Forcing leads on longer clips, shows the smallest drop from 5s to 60s among methods reporting all four lengths, and preserves subjects and scenes through leave-and-return. The same design scales to Wan2.2 5B, producing more physically plausible, realistic, and dynamic videos and, to our knowledge, the first public 5B model on this forcing line.

Jiaming Zhang, Xinyu Wang, Huafeng Shi +2

State Key Laboratory for Novel Software Technology, Nanjing University · Tsinghua University · Kling Team, Kuaishou Technology +1

Streaming Video Generation Memory-Augmented Video Generation Autoregressive Video Generation Long-Horizon Video Generation

Ask PDF

Abstract

Video Generation

Explore similar work

Echo-Forcing: A Scene Memory Framework for Interactive Long Video Generation

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

Memory Forcing: Attendable Mid-Horizon History for Streaming Video Generation