# PackCache: A Training-Free Acceleration Method for Unified Autoregressive Video Generation via Compact KV-Cache

arXiv: https://arxiv.org/abs/2601.04359 · HTML: https://mlsanity.com/paper/2601.04359/

## Key facts

- Cited by 13 papers in the map (data through 14 Oct 2026).
- Placed in Fast autoregressive visual generation, one of 12 directions in Efficient LLM inference and serving.
- Builds on 52 of its 54 references that are in the map.
- First posted on arXiv 7 Jan 2026.

## Placed in

- Efficient LLM inference and serving / [Fast autoregressive visual generation](https://mlsanity.com/inference/fast-autoregressive-visual-generation/)

## Builds on (most cited first, 30 of 52)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://mlsanity.com/paper/1910.10683/) (2019)
- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [Qwen3 Technical Report](https://mlsanity.com/paper/2505.09388/) (2025)
- [PaLM: Scaling Language Modeling with Pathways](https://mlsanity.com/paper/2204.02311/) (2022)
- [Efficient Streaming Language Models with Attention Sinks](https://mlsanity.com/paper/2309.17453/) (2023)
- [VBench: Comprehensive Benchmark Suite for Video Generative Models](https://mlsanity.com/paper/2311.17982/) (2023)
- [GQA: Training Generalized Multi-Query Transformer Models from Multi-Head Checkpoints](https://mlsanity.com/paper/2305.13245/) (2023)
- [DeepSeek-V2: A Strong, Economical, and Efficient Mixture-of-Experts Language Model](https://mlsanity.com/paper/2405.04434/) (2024)
- [Kosmos-2: Grounding Multimodal Large Language Models to the World](https://mlsanity.com/paper/2306.14824/) (2023)
- [CogVideo: Large-scale Pretraining for Text-to-Video Generation via Transformers](https://mlsanity.com/paper/2205.15868/) (2022)
- [Chameleon: Mixed-Modal Early-Fusion Foundation Models](https://mlsanity.com/paper/2405.09818/) (2024)
- [SnapKV: LLM Knows What You are Looking for Before Generation](https://mlsanity.com/paper/2404.14469/) (2024)
- [DeepSeekMoE: Towards Ultimate Expert Specialization in Mixture-of-Experts Language Models](https://mlsanity.com/paper/2401.06066/) (2024)
- [Cosmos World Foundation Model Platform for Physical AI](https://mlsanity.com/paper/2501.03575/) (2025)
- [GLM-130B: An Open Bilingual Pre-trained Model](https://mlsanity.com/paper/2210.02414/) (2022)
- [H2O: Heavy-Hitter Oracle for Efficient Generative Inference of Large Language Models](https://mlsanity.com/paper/2306.14048/) (2023)
- [Fast Transformer Decoding: One Write-Head is All You Need](https://mlsanity.com/paper/1911.02150/) (2019)
- [VideoGPT: Video Generation using VQ-VAE and Transformers](https://mlsanity.com/paper/2104.10157/) (2021)
- [FlexGen: High-Throughput Generative Inference of Large Language Models with a Single GPU](https://mlsanity.com/paper/2303.06865/) (2023)
- [Quest: Query-Aware Sparsity for Efficient Long-Context LLM Inference](https://mlsanity.com/paper/2406.10774/) (2024)
- [VideoPoet: A Large Language Model for Zero-Shot Video Generation](https://mlsanity.com/paper/2312.14125/) (2023)
- [Phenaki: Variable Length Video Generation From Open Domain Textual Description](https://mlsanity.com/paper/2210.02399/) (2022)
- [Model Tells You What to Discard: Adaptive KV Cache Compression for LLMs](https://mlsanity.com/paper/2310.01801/) (2023)
- [From Slow Bidirectional to Fast Autoregressive Video Diffusion Models](https://mlsanity.com/paper/2412.07772/) (2024)
- [PyramidKV: Dynamic KV Cache Compression based on Pyramidal Information Funneling](https://mlsanity.com/paper/2406.02069/) (2024)
- [OmniQuant: Omnidirectionally Calibrated Quantization for Large Language Models](https://mlsanity.com/paper/2308.13137/) (2023)
- [Florence-2: Advancing a Unified Representation for a Variety of Vision Tasks](https://mlsanity.com/paper/2311.06242/) (2023)

## Built on by (most cited first, 13 of 13)

- [Relax Forcing: Relaxed KV-Memory for Consistent Long Video Generation](https://mlsanity.com/paper/2603.21366/) (2026)
- [Addressable Memory for Video World Models](https://mlsanity.com/paper/2608.07408/) (2026)
- [Future Forcing: Future-aware Training-free KV Cache Policy for Autoregressive Video Generation](https://mlsanity.com/paper/2605.30083/) (2026)
- [Efficient Video Diffusion Models: Advancements and Challenges](https://mlsanity.com/paper/2604.15911/) (2026)
- [Attend Locally, Remember Linearly: Linear Attention as Cross-Frame Memory for Autoregressive Video Diffusion](https://mlsanity.com/paper/2605.16579/) (2026)
- [FastSTAR: Spatiotemporal Token Pruning for Efficient Autoregressive Video Synthesis](https://mlsanity.com/paper/2603.07192/) (2026)
- [CIVIC: End-to-End Sequence Compactness for Efficient Vision-Language Models](https://mlsanity.com/paper/2605.28115/) (2026)
- [SparSTAR: Sparse Attention for SpaceTime AutoRegressive Video Synthesis](https://mlsanity.com/paper/2608.10519/) (2026)
- [DeCoPrune: Efficient KV-Cache Pruning for Autoregressive Video Diffusion via Denoising Consistency](https://mlsanity.com/paper/2609.39096/) (2026)
- [MosaiChunk: Compositing Spatio-Temporal Memory for Autoregressive Video Generation](https://mlsanity.com/paper/2610.02153/) (2026)
- [Keepsake: Selective Spatial Memory for Long-Horizon Video Generation](https://mlsanity.com/paper/2610.06588/) (2026)
- [CORAL-KV: Compact Cross-Layer Prediction and Residual Coding for Autoregressive Video KV Caches](https://mlsanity.com/paper/2610.12514/) (2026)
- [TempoKV: Predictive KV Cache Quantization for Autoregressive Video Generation](https://mlsanity.com/paper/2610.15324/) (2026)

## Abstract (arXiv)

A unified autoregressive model is a Transformer-based framework that addresses diverse multimodal tasks (e.g., text, image, video) as a single sequence modeling problem under a shared token space. Such models rely on the KV-cache mechanism to reduce attention computation from O(T^2) to O(T); however, KV-cache size grows linearly with the number of generated tokens, and it rapidly becomes the dominant bottleneck limiting inference efficiency and generative length. Unified autoregressive video generation inherits this limitation. Our analysis reveals that KV-cache tokens exhibit distinct spatiotemporal properties: (i) text and conditioning-image tokens act as persistent semantic anchors that consistently receive high attention, and (ii) attention to previous frames naturally decays with temporal distance. Leveraging these observations, we introduce PackCache, a training-free KV-cache management method that dynamically compacts the KV cache through three coordinated mechanisms: condition anchoring that preserves semantic references, cross-frame decay modeling that allocates cache budget according to temporal distance, and spatially preserving position embedding that maintains coherent 3D structure under cache removal. In terms of efficiency, PackCache accelerates end-to-end generation by 1.7-2.2x on 48-frame long sequences, showcasing its strong potential for enabling longer-sequence video generation. Notably, the final four frames - the portion most impacted by the progressively expanding KV-cache and thus the most expensive segment of the clip - PackCache delivers a 2.6x and 3.7x acceleration on A40 and H200, respectively, for 48-frame videos.