C3M is a cross‑session multimodal memory system designed for long‑horizon tasks that must preserve and retrieve evidence across sessions within a limited, query‑blind memory budget. It maintains a bounded active index of source text‑image evidence, using relation‑aware updates to keep safe redundancy while preserving complementary and incompatible records. At query time, budgeted routing selects useful index pages and expands their associated source evidence under a fixed reader budget, creating a compact, provenance‑preserving memory that retains temporal distinctions and source links for reliable downstream reasoning.
By Xueshu Chen, Yan Wang, Zihao Xue, Jiefu Li, Zhenfang Liu, Jayden Chen, Zhen Bi, Jungang Lou
arXiv:2607. 28678v1 Announce Type: new Abstract: Multimodal agents operating in long-horizon environments must build and continually update multimedia memories to support entity-consistent, temporally grounded reasoning.
By Xinkui Zhao, Enbo Chen, Yifan Zhang, Chang Liu, Guanjie Cheng, Naibo Wang, Yueshen Xu
GraphMemix introduces a combinatorial‑optimization graph memory framework that constructs query‑aware evidence forests for long‑term multimodal agent memory. It expands seed memories via schema and semantic relations, decouples memory support from relation verification to reduce redundancy, and optimizes a forest‑format context within a maximum evidence budget. Experiments on four benchmarks show significant accuracy gains and a new Pareto frontier between accuracy and lifecycle cost.
By Geng Li, Yuhao Wang, Dong Li, Jianye Hao, Yuxin Peng
GraphMemix introduces a combinatorial‑optimization graph memory framework that organizes long‑term multimodal agent memory as query‑aware evidence forests. It constructs candidate graphs by expanding seed memories through schema and semantic relations, then decouples evidence utility from anchor‑conditioned relation verification to reduce redundancy, and finally optimizes a forest‑format memory context within a maximum evidence budget. Experiments on four benchmarks show significant accuracy gains and a new Pareto frontier between accuracy and lifecycle cost.
arXiv:2607. 29440v1 Announce Type: new Abstract: Long-term multimodal memory must support not only retrieving relevant information but also computing over observations accumulated across interactions.
By Zhoujin Tian, Yao Tian, Hao Zhang, Cheng Chen, Yakun Li, Lei Zhang, Xiaofang Zhou
EdgeMem is a new agent-memory method that preserves original interaction turns and organizes them using complementary content, temporal, and episodic cues via a multi‑anchor hypergraph. It performs lightweight local processing, returning source evidence directly and reserving LLM use only for final answer generation. Experiments on LoCoMo and LongMemEval‑S demonstrate strong retrieval and memory‑grounded question answering, with EdgeMem achieving the highest strict‑judge score among seven systems on LoCoMo while requiring no generative‑LLM calls for construction and retrieval.
By Zeyang Cui, Jiannong Cao, Zhiyuan Wen, Bo Yuan, Junlan Feng, Shengyuan Chen
An agent that interacts with users over long periods must recall facts, preferences, events, and changes from a continuously growing interaction history. Existing memory systems often compress interac...
EnSIMem is an entity‑structured long‑term memory architecture designed for agents that interact with users over extended periods. It organizes interactions into theme‑coherent episodes and creates dialogue‑grounded index entries of the form [entity][entity type][property:value], preserving source turns, temporal data, and multimodal fields. During online interaction, the agent decomposes requests into evidence requirements, performs entity‑property lookup, and retrieves the necessary evidence to generate responses directly from preserved source material rather than lossy summaries.
By Xuanyu Meng, Xing Fan, Xinyi Fan, Chenlei Guo, Yixuan Xie, Jiawei Han
The paper introduces ViMoD, a lightweight framework that dynamically manages visual memory for multimodal large language models. ViMoD uses Deformable Aggregation of Region-wise Tokens (DART) to create content‑adaptive coarse representations linked to recoverable fine tokens, and Temporal Routing for Adaptive Contextual Evidence (TRACE) to anticipate and adjust evidence needs during reasoning. On the Qwen3‑VL‑4B model, ViMoD achieves superior performance across eight reasoning benchmarks with only a 20% visual token budget and minimal additional parameters.
By Yicheng Xue, Han Wu, Jufeng Yang, Minjing Dong, Xinghao Chen, Hanting Chen, Jianyuan Guo
MemFuseBench is a benchmark for multi‑source memory fusion that generates source‑tagged observations, evidence‑grounded questions, and adversarial distractors using a Scene‑to‑Sensor pipeline. The proposed MemFuse system preserves source‑level evidence in atomic memory and clusters related events into fused memory via a causal fusion graph, enabling traceable retrieval of dispersed observations. Experiments show that MemFuse outperforms other memory systems across all LLM settings, especially on questions requiring cross‑source evidence fusion.
By Chao Li, Yuanfa Li, Wenhao Wu, Xule Liu, Zhi Wang, Kun Shao
arXiv:2606. 29788v1 Announce Type: new Abstract: When a multimodal AI agent is asked to forget a fact, current memory systems usually delete the text entry and report success.
By Kuan Wang, Chao Zhang
The paper introduces AutoMMemo, a framework that lets agents automatically design memory mechanisms for learning from multimodal experience. Each memory mechanism is represented as an executable memo program that specifies how past episodes are stored, organized, and retrieved. Through iterative evaluation, reflection-guided mutation, and budget-aware tree search, AutoMMemo learns effective memory designs that improve agent performance on GUI/Web navigation and multimodal visual reasoning tasks while keeping token and interaction costs reasonable.
By Xingyu Sui, Weixiang Zhao, Yongxin Tang, Yanyan Zhao, Yang Wu, Dandan Tu, Bing Qin