Large language model agents have shown strong capabilities in generating coherent and contextually appropriate responses, yet robust long-horizon dialogue remains limited by the lack of external memory that is traceable, updatable, and diagnostically transparent. Existing memory-augmented agents often store memories as isolated records or overwritable states, making it difficult to preserve how information originates, evolves, conflicts, or becomes obsolete over time.
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
arXiv:2607. 26455v1 Announce Type: cross Abstract: Large language models (LLMs) have demonstrated strong capabilities in knowledge acquisition and reasoning, yet their ability to retain previously acquired knowledge under repeated updates remains insufficiently understood.
By Ruxi Gu, Zhenliang Zhang, Wei Wang
arXiv:2606. 05684v1 Announce Type: new Abstract: A central challenge for language agents is utilizing past experience to adapt to dynamic test-time conditions.
By Yunxiang Zhang, Yiheng Li, Ali Payani, Lu Wang
arXiv:2606. 07909v2 Announce Type: replace Abstract: Modern large language model (LLM) agents can use external tools to help users solve complex tasks.
By Suleyman Armagan Er, Danilo Ribeiro, Yogesh Virkar, Surafel Lakew, Adi Kalyanpur, James Gung, Thomas Delteil, Arshit Gupta
The paper introduces MERIT, a benchmark that evaluates the marginal benefit of long‑term memory for tool‑using large language model agents while explicitly accounting for cost. MERIT provides episodic tool‑use tasks across three domains, verifies dependence on earlier‑episode facts, and measures memory operations in tokens and dollars. Experiments on GPT‑4.1‑mini, Claude Haiku 4.5, and Claude Sonnet 5 show that memory can significantly improve task success, but its utility varies widely across models and memory implementations, and full replay is rarely cost‑effective.
By Shweta Mishra, Shashank Mishra
arXiv:2607. 04089v1 Announce Type: new Abstract: Lifelong agents need more than larger context windows and better retrieval.
By Sukanta Ganguly
Large language models (LLMs) have demonstrated strong capabilities in knowledge acquisition and reasoning, yet their ability to retain previously acquired knowledge under repeated updates remains insufficiently understood. Existing evaluation paradigms primarily focus on single-step reasoning or static knowledge editing, which fail to capture the temporal dynamics of knowledge retention and degradation during continual model modification.
The paper introduces RD-Forget, a training‑free framework that separates what a persistent language agent stores from what it uses at answer time. It keeps a source archive of all observations while a query‑conditioned memory view filters evidence relevant to the current question, using a frozen language‑model curator to group facts into semantic slots and preserve multi‑hop relations. The approach employs rate‑distortion principles to stay within a memory budget and demonstrates improvements across conversational memory, knowledge updating, fact consolidation, long‑context reasoning, and personalization tasks.
By Yuhang Li, Yuchen Li
arXiv:2606. 06787v1 Announce Type: new Abstract: Large Language Models (LLMs) show promise as tool-using agents but remain limited in long-horizon tasks that require remembering, organizing, and reusing knowledge.
By Runzhe Wang, Huilin Lu, Shengjie Liu, Li Dong, Jason Zhu
The paper introduces AMA, a framework that uses multiple agents—Constructor, Retriever, Judge, and Refresher—to manage memory for large language model agents. AMA’s hierarchical memory design dynamically adjusts retrieval granularity to match task complexity, while the Judge and Refresher ensure relevance, consistency, and timely updates. Experiments on long-context benchmarks show AMA outperforms existing baselines and cuts token usage by about 80% compared to full-context approaches.
By Weiquan Huang, Zixuan Wang, Hehai Lin, Sudong Wang, Bo Xu, Qian Li, Beier Zhu, Linyi Yang, Chengwei Qin
MemoryArena is a new evaluation gym that benchmarks agent memory in interdependent multi‑session tasks. Unlike prior benchmarks that test memorization or single‑session action in isolation, MemoryArena requires agents to acquire memory while interacting with the environment and then use that memory to guide future decisions across a range of tasks such as web navigation, planning, information search, and formal reasoning. The benchmark reveals that agents excelling on existing long‑context memory tests perform poorly here, highlighting a gap in current memory evaluation methods.
By Zexue He, Yu Wang, Churan Zhi, Yuanzhe Hu, Tzu-Ping Chen, Lang Yin, Ze Chen, Tong Arthur Wu, Siru Ouyang, Zihan Wang, Jiaxin Pei, Julian McAuley, Yejin Choi, Alex Pentland