Multimodal Agent MemorySubmit a paper

AGMem

Naive Visual Memory is Not Enough: A Failure-Mode Study of GUI Agents

Seoyoung Choi, Minseok Ko, Hyunseok Lee, Kunwoong Kim, Woomin Song, Chanseok Jeon, et al. · 2026-06-12

Read the paperPDF

Management
T1 Static Memory
Memory is built once and reused without change during deployment.
Primary modality
GUI & computer use
Also: Image, text
Representation
Source- and modality-specific

How experience becomes memory

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

  1. action-grounded GUI crops

    1. Represented asRetained perceptual source evidenceSource- and modality-specific
    2. Stored inVector-indexed collection, Flat sequential records
    3. Retrieved bySimilarity search
    4. Source evidenceSource directly available
    AGMem stores local crops tied to successful actions or recoveries instead of full screenshots.Full paper: 3.2. Action-grounded visual memory · PDF pp. 5–6, 13–14 · Figure 1
    The memory bank is also an ordered collection of trajectory-segmented atomic memory steps, in addition to its vector index.Full paper: 3.2. Action-grounded visual memory · PDF pp. 5–6, 13–14 · Figure 1

Related systems