NeFut Logo NeFut
Admin Login

[CS.AI] GraphMemix: Query-Aware Evidence Forests for Long-Term Multimodal Agent Memory

Published at: 2026-08-30 22:00 Last updated: 2026-09-01 02:31
#AI #Machine Learning #Graph

Organizing long‑term memory for multimodal agents remains a challenging problem. Existing approaches either rely on expensive, query‑agnostic offline summaries or on naive embedding‑similarity matching, which yields incomplete and redundant context. To tackle these issues we introduce GraphMemix, a combinatorial‑optimization graph‑memory framework that treats memory organization as a query‑aware evidence‑forest construction.

The method consists of three key components:

  1. Candidate graph construction – expands multi‑view seed memories via schema and semantic relations to obtain query‑relevant original context.
  2. Evidence utility and activation cost – decouples direct memory support from anchor‑conditioned relation verification, suppressing redundant or conflicting information.
  3. Forest optimization – jointly selects a forest‑structured memory subgraph under a maximum evidence budget while preserving a reliable relational structure.

By structuring memory as a query‑relevant subgraph, the approach avoids substantial lifecycle overhead and recovers low‑similarity complementary evidence. Experiments on four long‑term multimodal memory benchmarks show significant gains across various foundation models and establish a new Pareto frontier between accuracy and lifecycle cost.

Blogger's Review: GraphMemix achieves impressive accuracy while markedly reducing retrieval cost, highlighting the power of query‑aware memory organization.

Original Source: https://arxiv.org/abs/2608.26983

[h] Back to Home