Multimodal Agent MemorySubmit a paper

GenEvolve

GenEvolve: Self-Evolving Image Generation Agents via Tool-Orchestrated Visual Experience Distillation

Sixiang Chen, Zhaohu Xing, Tian Ye, Xinyu Geng, Yunlong Lin, Jianyu Lai, et al. · 2026-05-22

Management
T1 Static Memory
Memory is built once and reused without change during deployment.
Primary modality
Image
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. structured visual experience distilled from best-worst trajectory differences

    1. Represented asSpatial, temporal & cross-modal structured memoryCross-represented explicit
    2. Stored inFlat sequential records
    3. Retrieved byPlanner / tool-assisted retrieval
    4. Source evidenceSource not retained
    Structured visual-experience entries aggregating successful and failed generation patterns. agent consults distilled visual experience to guide subsequent generations.Full-paper method / memory representation and retrieval · PDF p. 3 · Figure 1 / method overview

Related systems