CogMem: A Cognitive Memory Architecture for Sustained Multi-Turn Reasoning in Large Language Models

TL;DR

CogMem introduces a layered, memory-augmented architecture for large language models that enhances multi-turn reasoning, reduces context issues, and improves coherence and reliability over extended interactions.

Contribution

The paper presents CogMem, a novel cognitive-inspired memory architecture with three layers that supports sustained reasoning and addresses limitations of current LLMs in multi-turn dialogues.

Findings

Mitigates reasoning failures and hallucinations.

Controls context growth effectively.

Enhances reasoning consistency over multiple turns.

Abstract

Large language models (LLMs) excel at single-turn reasoning but often lose accuracy and coherence over extended, multi-turn interactions. Recent evaluations such as TurnBench highlight recurring failure modes-reasoning bias, task drift, hallucination, overconfidence, and memory decay. Current approaches typically append full conversational histories, causing unbounded context growth, higher computational costs, and degraded reasoning efficiency. We introduce CogMem, a cognitively inspired, memory-augmented LLM architecture that supports sustained iterative reasoning through structured, persistent memory. CogMem incorporates three layers: a Long-Term Memory (LTM) that consolidates cross-session reasoning strategies; a Direct Access (DA) memory that maintains session-level notes and retrieves relevant long-term memories; and a Focus of Attention (FoA) mechanism that dynamically reconstructs concise, task-relevant context at each turn. Experiments on TurnBench show that this layered design mitigates reasoning failures, controls context growth, and improves consistency across extended reasoning chains, moving toward more reliable, human-like reasoning in LLMs.