FadeMem: Biologically-Inspired Forgetting for Efficient Agent Memory
Authors:
- 3 May 2026
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 4011-4015
Abstract
Large language models deployed as autonomous agents face critical memory limitations, lacking selective forgetting mechanisms that lead to either catastrophic forgetting at context boundaries or information overload within them. While human memory naturally balances retention and forgetting through adaptive decay processes, current AI systems employ binary retention strategies that preserve everything or lose it entirely. We propose FadeMem, a biologically-inspired agent memory architecture that incorporates active forgetting mechanisms mirroring human cognitive efficiency. FadeMem implements differential decay rates across a dual-layer memory hierarchy, where retention is governed by adaptive exponential decay functions modulated by semantic relevance, access frequency, and temporal patterns. Through LLM-guided conflict resolution and intelligent memory fusion, our system consolidates related information while allowing irrelevant details to fade. Experiments on Multi-Session Chat, LoCoMo, and LTI-Bench demonstrate superior multi-hop reasoning and retrieval with 45% storage reduction, validating the effectiveness of biologically-inspired forgetting in agent memory systems.Keywords
All Related Versions
This publication has 20 references indexed in Scilit:
- Multimodal Multi-Agent Empowered Legal Judgment Prediction2026
- RAG-Driven Memory Architectures in Conversational LLMs—A Literature Review With Insights Into Emerging Agriculture Data SharingIEEE Access, 2025
- A Survey on the Memory Mechanism of Large Language Model based AgentsACM Transactions on Information Systems, 2025
- Machine Memory Intelligence: Inspired by Human Memory MechanismsEngineering, 2025
- Lost in the Middle: How Language Models Use Long ContextsTransactions of the Association for Computational Linguistics, 2024
- A biologically inspired architecture with switching units can learn to generalize across backgroundsNeural Networks, 2023
- Bio-inspired cognitive architecture of episodic memoryCognitive Systems Research, 2022
- Biologically Inspired Sleep Algorithm for Reducing Catastrophic Forgetting in Neural NetworksProceedings of the AAAI Conference on Artificial Intelligence, 2020
- Overcoming catastrophic forgetting in neural networksProceedings of the National Academy of Sciences, 2017
- The Psychology and Neuroscience of ForgettingAnnual Review of Psychology, 2004