Multimodal Agent MemorySubmit a paper

Ptah

Towards Verifiable Multimodal Deep Research: A Multi-Agent Harness for Interleaved Report Generation

Chenghao Zhang, Guanting Dong, Yufan Liu, Tong Zhao, Xiaoxi Li, Zhicheng Dou · 2026-06-03

Read the paperPDF

Management
T2 Dynamic Memory
Memory changes during deployment through a stable management policy.
Primary modality
Image
Also: 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. source-aligned images in a Visual Working Memory

    1. Represented asRetained perceptual source evidenceSource- and modality-specific
    2. Stored inFlat sequential records
    3. Retrieved byPlanner / tool-assisted retrieval
    4. Source evidenceSource directly available
    Visual Working Memory entries (source-aligned images) plus claim-grounded evidence records. declarative multimodal tool use over Visual Working Memory during report composition.Full-paper method / memory representation and retrieval · PDF pp. 3–4 · Figure 1 / method overview

How memory changes (T2 operations)

  • Update · Rule-based update

    During operation, Ptah applies fixed-rule insertion or replacement to source-aligned images in a Visual Working Memory; the memory state changes while the governing update policy remains fixed.

Related systems