A Simple Yet Strong Baseline for Long-Term Conversational Memory of LLM Agents

• URL: http://arxiv.org/abs/2511.17208v1
• Date: Fri, 21 Nov 2025 12:41:17 GMT
• Title: A Simple Yet Strong Baseline for Long-Term Conversational Memory of LLM Agents
• Authors: Sizhe Zhou,
• Abstract summary: LLM-based conversational agents still struggle to maintain coherent, personalized interaction over many sessions.<n>Motivated by neo-Davidsonian event semantics, we propose an event-centric alternative that represents conversational history as short, event-like propositions.<n>Our design aims to preserve information in a non-compressive form and make it more accessible, rather than more lossy.
• Score: 4.706565675142042
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: LLM-based conversational agents still struggle to maintain coherent, personalized interaction over many sessions: fixed context windows limit how much history can be kept in view, and most external memory approaches trade off between coarse retrieval over large chunks and fine-grained but fragmented views of the dialogue. Motivated by neo-Davidsonian event semantics, we propose an event-centric alternative that represents conversational history as short, event-like propositions which bundle together participants, temporal cues, and minimal local context, rather than as independent relation triples or opaque summaries. In contrast to work that aggressively compresses or forgets past content, our design aims to preserve information in a non-compressive form and make it more accessible, rather than more lossy. Concretely, we instruct an LLM to decompose each session into enriched elementary discourse units (EDUs) -- self-contained statements with normalized entities and source turn attributions -- and organize sessions, EDUs, and their arguments in a heterogeneous graph that supports associative recall. On top of this representation we build two simple retrieval-based variants that use dense similarity search and LLM filtering, with an optional graph-based propagation step to connect and aggregate evidence across related EDUs. Experiments on the LoCoMo and LongMemEval$_S$ benchmarks show that these event-centric memories match or surpass strong baselines, while operating with much shorter QA contexts. Our results suggest that structurally simple, event-level memory provides a principled and practical foundation for long-horizon conversational agents. Our code and data will be released at https://github.com/KevinSRR/EMem.

Related papers

• Event Extraction in Large Language Model [99.94321497574805]
  We argue that EE should be viewed as a system component that provides a cognitive scaffold for LLM centered solutions.<n>This survey covers EE in text and multimodal settings, organizing tasks and taxonomy, tracing method evolution from rule based and neural models to instruction driven and generative frameworks.
  arXiv  Detail & Related papers  (2025-12-22T16:22:14Z)
• Beyond Fact Retrieval: Episodic Memory for RAG with Generative Semantic Workspaces [5.110309385104824]
  Large Language Models (LLMs) face fundamental challenges in long-context reasoning.<n>Current solutions fail to build the space-time-anchored narrative representations required for tracking entities through episodic events.<n>We propose the textbfGenerative Semantic Workspace (GSW), a neuro-inspired generative memory framework.
  arXiv  Detail & Related papers  (2025-11-10T19:59:34Z)
• SGMem: Sentence Graph Memory for Long-Term Conversational Agents [14.89396085814917]
  We introduce SGMem (Sentence Graph Memory), which represents dialogue as sentence-level graphs within chunked units.<n>We show that SGMem consistently improves accuracy and outperforms strong baselines in long-term conversational question answering.
  arXiv  Detail & Related papers  (2025-09-25T14:21:44Z)
• From Single to Multi-Granularity: Toward Long-Term Memory Association and Selection of Conversational Agents [79.87304940020256]
  Large Language Models (LLMs) have been widely adopted in conversational agents.<n>MemGAS is a framework that enhances memory consolidation by constructing multi-granularity association, adaptive selection, and retrieval.<n> Experiments on four long-term memory benchmarks demonstrate that MemGAS outperforms state-of-the-art methods on both question answer and retrieval tasks.
  arXiv  Detail & Related papers  (2025-05-26T06:13:07Z)
• Toward Multi-Session Personalized Conversation: A Large-Scale Dataset and Hierarchical Tree Framework for Implicit Reasoning [30.54506564763053]
  We introduce ImplexConv, a large-scale long-term dataset with 2,500 examples, each containing approximately 100 conversation sessions.<n>We also propose TaciTree, a novel hierarchical tree framework that structures conversation history into multiple levels of summarization.
  arXiv  Detail & Related papers  (2025-03-10T07:59:41Z)
• FragRel: Exploiting Fragment-level Relations in the External Memory of Large Language Models [54.13671100638092]
  We propose a fragment-connected Hierarchical Memory based Large Language Models (LLMs) We formulate the fragment-level relations in external memory and present several instantiations for different text types. We validate the benefits of involving these relations on long story understanding, repository-level code generation, and long-term chatting.
  arXiv  Detail & Related papers  (2024-06-05T09:31:37Z)
• Walking Down the Memory Maze: Beyond Context Limit through Interactive Reading [63.93888816206071]
  We introduce MemWalker, a method that processes the long context into a tree of summary nodes. Upon receiving a query, the model navigates this tree in search of relevant information, and responds once it gathers sufficient information. We show that, beyond effective reading, MemWalker enhances explainability by highlighting the reasoning steps as it interactively reads the text; pinpointing the relevant text segments related to the query.
  arXiv  Detail & Related papers  (2023-10-08T06:18:14Z)
• Recursively Summarizing Enables Long-Term Dialogue Memory in Large Language Models [30.48902594738911]
  Given a long conversation, large language models (LLMs) fail to recall past information and tend to generate inconsistent responses.<n>We propose to generate summaries/ memory using large language models (LLMs) to enhance long-term memory ability.
  arXiv  Detail & Related papers  (2023-08-29T04:59:53Z)
• UniMC: A Unified Framework for Long-Term Memory Conversation via Relevance Representation Learning [15.313416157905685]
  We propose a Unified framework for Long-term Memory Conversations (UniMC) We decompose the main task into three subtasks based on probability graphs. Each subtask involves learning a representation for calculating the relevance between the query and memory.
  arXiv  Detail & Related papers  (2023-06-18T12:30:50Z)
• SCM: Enhancing Large Language Model with Self-Controlled Memory Framework [54.33686574304374]
  Large Language Models (LLMs) are constrained by their inability to process lengthy inputs, resulting in the loss of critical historical information.<n>We propose the Self-Controlled Memory (SCM) framework to enhance the ability of LLMs to maintain long-term memory and recall relevant information.
  arXiv  Detail & Related papers  (2023-04-26T07:25:31Z)
• History-Aware Hierarchical Transformer for Multi-session Open-domain Dialogue System [59.78425104243993]
  We propose History-Aware Hierarchical Transformer (HAHT) for multi-session open-domain dialogue. HAHT maintains a long-term memory of history conversations and utilizes history information to understand current conversation context. Experimental results on a large-scale Multi-Session Conversation dataset suggest that the proposed HAHT model consistently outperforms baseline models.
  arXiv  Detail & Related papers  (2023-02-02T06:54:33Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.