DreamBench‑SWE is a multi‑session benchmark designed to evaluate software agents’ memory hygiene, where later tasks rely on evidence from earlier sessions and are scored by executable hidden oracles. The study reports results from an original scaled v2 fold and a preregistered v2.1 successor audit, detailing performance across various memory configurations and conditions. The findings confirm that DreamBench‑SWE can discriminate between different memory setups, but do not establish superiority, equivalence, or broad generality of any external‑system mechanism.
By Sarthak Singh
arXiv:2608.21867v1 Announce Type: new
Abstract: LLM agents are moving from single-prompt use to long task streams in which reusable memory becomes a core capability for terminal, software-engineering...
By Haoyu Wang, Guangyuan Dong, He Liang, Zijing Zhang, Jiachen Luo, Chuang Liu, Chao Xue, Hao Tang
arXiv:2610.27714v1 Announce Type: new
Abstract: When does experience memory help a closed-loop LLM driving agent? Published gains cannot say, because the two factors we find to matter most, which war...
By Yingxin Jia, Xuanyi Li
The paper introduces a scale‑conditioned evaluation protocol for agent memory that measures how usable evidence degrades as irrelevant sessions accumulate. It logs memory trajectories and reports four diagnostics—budget‑compliant reliability, tail memory‑call burden, failure‑regime decomposition, and the usable‑scale boundary. Experiments on LongMemEval and LoCoMo show that reliability loss varies across memory interfaces and agents, highlighting that scalable‑memory claims must be conditioned on agent, interface, scale range, and interaction budget.
By Jiaqi Shao, Yiyi Lu, Yunzhen Zhang, Bing Luo
arXiv:2607. 27080v1 Announce Type: cross Abstract: Memory systems allow agents to retain and reuse information from past interactions, but they can also let malicious content persist.
By Xuanze Chen, Xukang Xie, Wentao Fu, Jiajun Zhou, Shanqing Yu, Qi Xuan
The paper introduces LSREP, a Longitudinal State‑Replay Evaluation Protocol designed to assess how conversational memory evolves over time, incorporating ordered replay, lifecycle schedules, repeated probes, evolving reference answers, and mechanism‑fidelity checks. It applies LSREP to ICE v2, a local‑first memory middleware, and reports that on three ordinary‑density datasets ICE v2 achieves near‑zero mean quality difference from vector‑RAG while using fewer fragments but slightly more prompt tokens, yet fails catastrophically on a dense dataset. In a public diagnostic, ICE v2 underperforms pure vector‑RAG on LongMemEval, revealing significant multi‑session and temporal failures and a quality‑cost trade‑off rather than superior efficiency.
By Deepesh Sonar
arXiv:2608. 12476v1 Announce Type: new Abstract: Long-term agent memory is usually treated as select--store--retrieve, but retrieval does not decide whether contradictory, superseded, retracted, deleted, or stale records may support an outgoing claim.
By Guodong Xu
arXiv:2608. 13883v1 Announce Type: new Abstract: Most agent-memory benchmarks test post-hoc recall, whereas MemoryArena evaluates whether memory supports interdependent, multi-session task completion.
By Chaoqun Zhan, Qiang Zhou, Guannan Li, Zhenqiang Huang, Qianjin Wang
arXiv:2608. 07429v1 Announce Type: new Abstract: Long-term memory enables language agents to reuse past facts, preferences, and task experience.
By Yan Zhou, Yue Ouyang, Kaiyang Zheng, Suncheng Xiang
The paper investigates how persistent memory in AI agents can lead to over‑trust in stale facts, creating a "Memory Trust Gap" that worsens as model capability increases. Using a benchmark with Benefit and Safety suites across Qwen3 models of varying sizes, the authors show that larger models are more prone to harmful over‑trust, especially when metadata is absent or misleading. They also demonstrate that mitigation strategies such as exposing metadata or pre‑resolving conflicts improve accuracy, but the effectiveness depends on model size and dataset.
By Jundong Hu, Shekar Ramachandran
arXiv:2608. 02613v1 Announce Type: cross Abstract: Edge-deployed personal memory assistants must handle private interpersonal conversations on-device with open-weight models.
By Jiadong Zhang, Xiaosong Ma
arXiv:2609.13149v1 Announce Type: new
Abstract: For local large language model agents, active context is a scarce resource: memory capacity, prefill latency, cache growth, and service objectives all...
By Aditya Karnam Gururaj Rao, Arjun Jaggi