Multimodal Agent MemorySubmit a paper

EvolvingAgent

EvolvingAgent: Curriculum Self-evolving Agent with Continual World Model for Long-Horizon Tasks

Tongtong Feng, Xin Wang, Zekai Zhou, Ren Wang, Yuwei Zhan, Guangyao Li, et al. · 2026-04-29

Read the paperPDF

Management
T2 Dynamic Memory
Memory changes during deployment through a stable management policy.
Primary modality
Embodied & 3D
Also: text
Representation
Cross-represented explicit

How experience becomes memory

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

  1. multimodal experience pool

    1. Represented asSpatial, temporal & cross-modal structured memoryCross-represented explicit
    2. Stored inFlat sequential records
    3. Retrieved byMetadata / spatiotemporal retrieval, Planner / tool-assisted retrieval
    4. Source evidenceSource not retained
    Continual World Model; multimodal experience updates; selected curriculum experiences. planner uses multimodal experiences and world knowledge to decompose tasks; controller uses world model for actions.Full-paper method / memory representation and retrieval · PDF p. 4 · Figure 1 / method overview

How memory changes (T2 operations)

  • Update · Optimization-based update

    During operation, EvolvingAgent applies a fixed operational optimization objective to multimodal experience pool; the memory state changes while the governing update policy remains fixed.

  • Consolidation · Experience abstraction

    Central contribution: closed-loop continual world-model updates select and integrate multimodal experiences. For EvolvingAgent, this specifically supports abstraction of episodes into reusable experience of multimodal experience pool.

Related systems