Towards Lifelong Dialogue Agents via Timeline-based Memory Management

2024-06-16

AI Summary

To address inefficient memory management, weak contextual consistency, and inadequate behavioral evolution modeling in long-term dialogues, this paper proposes a lifelong interactive dialogue agent framework. Methodologically: (1) it introduces the first memory timeline modeling paradigm, structuring historical memories as a temporal-causal graph and replacing passive forgetting with active, principled memory organization; (2) it designs a memory-timeline-driven retrieval-augmented generation mechanism; and (3) it establishes Counterfactual-driven TeaFarm, an evaluation framework enabling dynamic assessment of behavioral evolution. Experiments demonstrate significant improvements in multi-turn dialogue coherence and user behavioral modeling accuracy, outperforming state-of-the-art baselines on memory integration tasks. The work is accompanied by an open-source dataset and an interactive demonstration system.

Technology Category

• Machine Learning: Life-Long and Continual Learning
• Multiagent Systems: Adversarial Agents
• Cognitive Modeling & Cognitive Systems: Agent Architectures

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Studies of user behavior, including longitudinal effects of personalized systems
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

To achieve lifelong human-agent interaction, dialogue agents need to constantly memorize perceived information and properly retrieve it for response generation (RG). While prior studies focus on getting rid of outdated memories to improve retrieval quality, we argue that such memories provide rich, important contextual cues for RG (e.g., changes in user behaviors) in long-term conversations. We present THEANINE, a framework for LLM-based lifelong dialogue agents. THEANINE discards memory removal and manages large-scale memories by linking them based on their temporal and cause-effect relation. Enabled by this linking structure, THEANINE augments RG with memory timelines - series of memories representing the evolution or causality of relevant past events. Along with THEANINE, we introduce TeaFarm, a counterfactual-driven evaluation scheme, addressing the limitation of G-Eval and human efforts when assessing agent performance in integrating past memories into RG. A supplementary video for THEANINE and data for TeaFarm are at https://huggingface.co/spaces/ResearcherScholar/Theanine.

Problem

• Dialogue Systems
• Memory Management
• Context Understanding

Innovation

• THEANINE
• memory utilization
• conversational AI