Multimodal Agent MemorySubmit a paper

MemVerse

MemVerse: Multimodal Memory for Lifelong Learning Agents

Junming Liu, Yifei Sun, Weihua Cheng, Haodong Lei, Yirong Chen, Licheng Wen, et al. · 2025-12-03

Read the paperPDF

Management
T2 Dynamic Memory
Memory changes during deployment through a stable management policy.
Primary modality
Image
Also: text
Representation
Hybrid
Distinct operative representation paths are explicitly linked for downstream retrieval, reasoning, or control.

How experience becomes memory

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

  1. hierarchical multimodal knowledge graph

    1. Represented asSpatial, temporal & cross-modal structured memoryCross-represented explicit
    2. Stored inGraph / relational store, Hierarchical multi-tier store
    3. Retrieved byStructure traversal, Similarity search
    4. Source evidenceSource linked & recoverable
    Raw multimodal experience becomes hierarchical long-term knowledge.Full paper: Method · PDF p. 8 · Figure 1

  2. distilled parametric memory

    1. Represented asExperience-conditioned parametric memoryInternalized latent & parametric
    2. Stored inModular adapter bank
    3. Retrieved byLatent / parametric activation
    4. Source evidenceSource not retained
    Periodic distillation transfers essential knowledge into parameters.Full paper: Without Memory · PDF p. 2 · Figure 1

How memory changes (T2 operations)

  • Consolidation · Experience abstraction

    A fixed pipeline abstracts experience into hierarchical knowledge. For MemVerse, this specifically supports abstraction of episodes into reusable experience of hierarchical multimodal knowledge graph.

  • Update · Optimization-based update

    Periodic distillation writes external knowledge into parameters. For MemVerse, this specifically supports a fixed operational optimization objective of hierarchical multimodal knowledge graph.

  • Forgetting · Progressive decay

    Adaptive forgetting bounds memory growth under a fixed rule. For MemVerse, this specifically supports progressive decay of stored importance or retention of hierarchical multimodal knowledge graph.

Related systems