← Back to all news
arXiv AI By Wenjun Wang, Yuchen Fang, Fengrui Liu, Zibo Liang, Kai Zheng

Experience Memory Graph: Extracting Reusable Correction Knowledge from Agent Trajectories

Read the original on arXiv AI →

The paper introduces Experience Memory Graph (EMG), a framework that transforms agent failure recovery into a graph matching problem. By converting failed and successful trajectories into directed action decision graphs, EMG extracts common subgraphs and edit paths that reveal explicit correction strategies, storing them in a memory graph with intra‑task nodes and cross‑task edges. During testing, EMG retrieves relevant insights to guide the agent in a single attempt, improving performance on ALFWorld and ScienceWorld without extra reflection rounds.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.

Related stories

arXiv AI
Oct 1

Agent Error Dataset: Scaling 50,000 Error--Diagnosis Pairs for Failure Analysis and Error-Aware Post-Training

The Agent Error Dataset (AED) presents 50,228 error–diagnosis pairs collected from 9,961 source tasks across 33 environments, 19 harness families, and 23 policy models in text‑based agent systems. A five‑stage Agentic Error‑to‑Training (AET) pipeline generates diagnoses and proposed corrections, verifies them against recorded evidence, and creates separate training views for diagnosis and actor recovery. Experiments show that first‑proposal corrections improve verifier pass rates from 18.4% to 51.1%, and fine‑tuning with full‑diagnosis data raises Qwen3‑8B’s exact‑step agreement from 47.2% to 63.6% on a holdout set.

By Kunlun Zhu, Xuyan Ye, Yibo Li, Cheng Qian, Beibin Li, Heng Ji
llmsagentsfine-tuning
More like this →
arXiv AI
Oct 16

Beyond Downstream QA: Rethinking How We Evaluate Agent Memory

The paper critiques current agent memory benchmarks that rely on downstream QA accuracy, arguing they conflate memory correctness with retrieval and generation quality, leading to misleading results. It introduces FEB-30, a new benchmark of 30 fact‑evolution trajectories across six diagnostic categories and three difficulty levels, generated automatically to provide deterministic ground truth. Using FEB-30, the authors show a significant divergence between intrinsic and extrinsic evaluations, revealing that a representative memory system correctly classifies memory operations only 20% of the time—an error invisible to traditional QA metrics.

By Sebastian-Antonio Toma, Jean de Kernier, Mike Ply, Oliviu Matei
agentsbenchmarks
More like this →
arXiv AI
Jun 30

Evaluating Memory in LLM Agents via Incremental Multi-Turn Interactions

arXiv:2507. 05257v4 Announce Type: replace-cross Abstract: Recent benchmarks for Large Language Model (LLM) agents primarily focus on evaluating reasoning, planning, and execution capabilities, while another critical component-memory, encompassing how agents memorize, update, and retrieve long-term information-is under-evaluated due to the lack of benchmarks.

By Yuanzhe Hu, Yu Wang, Julian McAuley
llmsragagentsbenchmarks
More like this →