Multimodal Agent MemorySubmit a paper

MovieDreamer

MovieDreamer: Hierarchical Generation for Coherent Long Visual Sequence

Canyu Zhao, Mingyu Liu, Wen Wang, Weihua Chen, Fan Wang, Hao Chen, et al. · 2025-11-25

Read the paperPDF

Management
T1 Static Memory
Memory is built once and reused without change during deployment.
Primary modality
Video & streaming
Also: text
Representation
Hybrid
Distinct operative representation paths are deliberately linked or jointly consumed by downstream reasoning, generation, or control.

How experience becomes memory

Each path is one memory component, traced from what is stored to how it is found again.

  1. Compact keyframe tokens within the long context

    1. Represented asRetained perceptual source evidenceSource- and modality-specific
    2. Stored inHierarchical multi-tier store
    3. Retrieved byStructure traversal
    4. Source evidenceSource directly available
    Autoregressive keyframe tokens (~2 per keyframe); Diffusion-decoded keyframes. Long videos are extended by anchoring the continuation on the previous first/last frame within the AR context.Full-paper method / memory representation and retrieval · PDF p. 25 · Figure 1 / method overview

  2. Compact keyframe tokens within the long context

    1. Represented asModality-specific encodings & geometric stateSource- and modality-specific
    2. Stored inHierarchical multi-tier store
    3. Retrieved byStructure traversal
    4. Source evidenceRetained but not accessible to the agent
    Autoregressive keyframe tokens (~2 per keyframe); Diffusion-decoded keyframes. Long videos are extended by anchoring the continuation on the previous first/last frame within the AR context.Full-paper method / memory representation and retrieval · PDF p. 25 · Figure 1 / method overview

Related systems