FadeMem: Biologically-Inspired Forgetting for Efficient Agent Memory

TL;DR

FadeMem introduces a biologically-inspired memory system for autonomous agents that employs adaptive decay and selective forgetting, significantly reducing storage needs while improving reasoning and retrieval capabilities.

Contribution

The paper presents FadeMem, a novel memory architecture that mimics human forgetting through differential decay, enhancing efficiency and performance in AI agents.

Findings

Achieves 45% storage reduction

Improves multi-hop reasoning and retrieval

Demonstrates effectiveness across multiple benchmarks

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.