Oblivion: Self-Adaptive Agentic Memory Control through Decay-Driven Activation

TL;DR

Oblivion introduces a decay-driven memory control framework for LLM agents, enabling dynamic forgetting and reinforcement to improve reasoning efficiency and context management.

Contribution

It presents a novel decay-based approach to memory control that decouples read and write paths, enhancing hierarchical memory organization in LLM agents.

Findings

Oblivion dynamically balances memory access and reinforcement.

It improves reasoning efficiency in long-horizon tasks.

The framework adapts to shifting contexts effectively.

Abstract

Human memory adapts through selective forgetting: experiences become less accessible over time but can be reactivated by reinforcement or contextual cues. In contrast, memory-augmented LLM agents rely on "always-on" retrieval and "flat" memory storage, causing high interference and latency as histories grow. We introduce Oblivion, a memory control framework that casts forgetting as decay-driven reductions in accessibility, not explicit deletion. Oblivion decouples memory control into read and write paths. The read path decides when to consult memory, based on agent uncertainty and memory buffer sufficiency, avoiding redundant always-on access. The write path decides what to strengthen, by reinforcing memories contributing to forming the response. Together, this enables hierarchical memory organization that maintains persistent high-level strategies while dynamically loading details as needed. We evaluate on both static and dynamic long-horizon interaction benchmarks. Results show that Oblivion dynamically adapts memory access and reinforcement, balancing learning and forgetting under shifting contexts, highlighting that memory control is essential for effective LLM-agentic reasoning.