Rethinking Memory as Continuously Evolving Connectivity
2026-05-27 • Computation and Language
Computation and LanguageArtificial IntelligenceMachine LearningMultiagent SystemsMultimedia
AI summaryⓘ
The authors found that current AI agents remember things in ways that don't change easily, which makes them less useful in changing situations. They created FluxMem, a new memory system that treats memory like a graph with connected points and updates these connections over time based on feedback. This helps the AI fix mistakes, focus on important details, and remember useful steps for later. They tested FluxMem on three different tasks and saw it worked better than older methods. The authors will share their code online for others to use.
memory-augmented LLMheterogeneous graphfeedback-driven refinementmemory generalizabilityagentic environmentstopology evolutionprocedural circuitslong-term consolidation
Authors
Jizhan Fang, Buqiang Xu, Zhixian Wang, Haoliang Cao, Xinle Deng, Baohua Dong, Hangcheng Zhu, Ruohui Huang, Gang Yu, Ying Wei, Guozhou Zheng, Feiyu Xiong, Haofen Wang, Huajun Chen, Ningyu Zhang
Abstract
Existing memory-augmented LLM agents often treat memory as a static repository with pre-defined representations and fixed retrieval pipelines, which is brittle in dynamic agentic environments where feedback, task variation, and heterogeneous signals continuously reshape what should be remembered and how it should be connected. To address this, we propose FluxMem, a connectivity-evolving memory framework that models memory as a heterogeneous graph and progressively refines its topology through three stages: initial connection formation, feedback-driven refinement, and long-term consolidation. During execution, FluxMem repairs missing links, prunes interference, aligns abstraction granularity, and distills recurrent successful trajectories into reusable procedural circuits, guided by one metric for memory generalizability and evolutionary maturity. Across three fundamentally distinct benchmarks including LoCoMo, Mind2Web, and GAIA, FluxMem achieves consistent state-of-the-art performance, demonstrating strong adaptation and generalization in complex agentic environments. The code will be open-sourced in https://github.com/zjunlp/LightMem.