Compress to Remember: Learning Compact Memory via On-Policy Distillation for Long Video Generation

arXiv:2609.36364 · cs.CV, cs.AI, cs.LG · Submitted 2026-09-28 · Read on arXiv

cs.CV, cs.AI, cs.LG

Submitted: 2026-09-28

Updated: 2026-09-28

Terminology

• Distillation — 18× in this paper · explained in 2 episodes
• Consistency — 16× in this paper · explained in 3 episodes
• Benchmark — 10× in this paper · explained in 2 episodes
• On-Policy Distillation — 8× in this paper · explained in 3 episodes
• Adaptation — 6× in this paper · explained in 1 episode
• Embeddings — 6× in this paper · explained in 2 episodes
• Prediction-Aligned Context Compaction — 6× in this paper
• Transformer — 6× in this paper · explained in 2 episodes
• Ablation — 4× in this paper · explained in 1 episode
• Generative Model — 3× in this paper · explained in 2 episodes
• Generative Models — 3× in this paper · explained in 2 episodes
• Redundancy — 3× in this paper · explained in 2 episodes
• World Model — 3× in this paper · explained in 7 episodes
• AI Agents — 1× in this paper · explained in 2 episodes
• Ablation study — 1× in this paper · explained in 1 episode
• Adam — 1× in this paper · explained in 2 episodes
• Aggregation — 1× in this paper · explained in 1 episode
• Alignment — 1× in this paper · explained in 1 episode
• Allocator — 1× in this paper · explained in 1 episode
• Consistency Distillation — 1× in this paper · explained in 2 episodes
• Foundation Model — 1× in this paper · explained in 3 episodes
• Foundation Models — 1× in this paper · explained in 7 episodes
• Gradient Flow — 1× in this paper · explained in 2 episodes
• Grounding — 1× in this paper · explained in 3 episodes
• KV cache — 1× in this paper · explained in 3 episodes
• Optical Flow — 1× in this paper · explained in 2 episodes
• RAG — 1× in this paper · explained in 3 episodes
• Scaling Laws — 1× in this paper · explained in 4 episodes

Sources

• Echo-Infinity: Learning Evolving Memory for Real-Time Infinite Video Generation
• Pyramid Forcing: Head-Aware Pyramid KV Cache Policy for High-Quality Long Video Generation
• SlotMemory: Object-Centric KV Memory for Streaming Long-Video Generation
• LongLive-RAG: A General Retrieval-Augmented Framework for Long Video Generation
• Scaling Laws for Neural Language Models
• MemRoPE: Training-Free Infinite Video Generation via Evolving Memory Tokens
• Rolling Sink: Bridging Limited-Horizon Training and Open-Ended Testing in Autoregressive Video Diffusion
• Long-Horizon Streaming Video Generation via Hybrid Attention with Decoupled Distillation
• PackForcing: Short Video Training Suffices for Long Video Sampling and Long Context Inference
• Movie Gen: A Cast of Media Foundation Models
• Wan: Open and Advanced Large-Scale Video Generative Models
• LongLive: Real-time Interactive Long Video Generation
• DySink: Dynamic Frame Sinks for Autoregressive Long Video Generation
• Deep Forcing: Training-Free Long Video Generation with Deep Sink and Participative Compression
• WorldKV: Efficient World Memory with World Retrieval and Compression
• MBench: A Comprehensive Benchmark on Memory Capability for Video World Models
• Relax Forcing: Relaxed KV-Memory for Consistent Long Video Generation
• Causal-rCM: A Unified Teacher-Forcing and Self-Forcing Open Recipe for Autoregressive Diffusion Distillation in Streaming Video Generation and Interactive World Models
• Causal Forcing: Autoregressive Diffusion Distillation Done Right for High-Quality Real-Time Interactive Video Generation

Related papers

• Loss Knows Best: Detecting Annotation Errors in Videos via Loss Trajectories
• AnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memories
• Benchmarking the Robustness of Foundation Models for Mammography under Domain Shift
• MambaX-Net: Dual-Input Mamba-Enhanced Cross-Attention Network for Longitudinal MRI Segmentation
• TeleOCR: Navigating Document Parsing Across Digital and Camera-Captured Documents
• A Survey on Efficient Vision-Language-Action Models