Multimodal Agent MemorySubmit a paper

DigimonGPT

DigimonGPT: An Evolvable Agent with Hierarchical Human-like Memory for Video Question Answering

Borui Li, Xingcai Zhang, Tianen Liu, Shuai Wang, Yun Cheng, Shuai Wang · 2026-03-14

Management
T2 Dynamic Memory
Memory changes during deployment through a stable management policy.
Primary modality
Video & streaming
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. intra-video declarative memory

    1. Represented asModality-specific encodings & geometric stateSource- and modality-specific
    2. Stored inHierarchical multi-tier store
    3. Retrieved bySimilarity search
    4. Source evidenceRetained but not accessible to the agent
    Video features and semantic contexts are retained in hierarchical declarative memory.Full paper: Memory · PDF p. 2 · Figure 1
  2. inter-task procedural memory

    1. Represented asSpatial, temporal & cross-modal structured memoryCross-represented explicit
    2. Stored inHierarchical multi-tier store
    3. Retrieved bySimilarity search, Planner / tool-assisted retrieval
    4. Source evidenceSource linked & recoverable
    Task-solving experiences are encoded as reusable procedural records and replayed by relevance and complexity.Full paper: Memory · PDF p. 2 · Figure 1

How memory changes (T2 operations)

  • Consolidation · Experience abstraction

    Historical question-answering episodes are abstracted into procedural experience. For DigimonGPT, this specifically supports abstraction of episodes into reusable experience of intra-video declarative memory.

  • Consolidation · Episodic / hierarchical summarization

    Hierarchical replay consolidates experience at multiple levels. For DigimonGPT, this specifically supports episodic or hierarchical summarization of intra-video declarative memory.

Related systems