Use and Disuse: Intent-Structured Experience Consolidation for Memory and Learning in LLM Agents
To address this, we propose Hippocam, a hierarchical memory and continual learning architecture.
ProofPaper ↗
Key points
- The evolution of Large Language Model agents from single-task execution to long-term autonomous operation highlights the critical challenge of transforming continuous experiences into reusable knowledge.
- Hippocam draws inspiration from two characteristics of human memory: cognitive processes selectively maintain information relevant to current goals, while long-term memories form gradually through repeated consolidation.
- Accordingly, Hippocam structures an agent's ongoing work as nested intents.
- Through this memory dynamic of use and disuse, Hippocam connects working context, long-term memory, knowledge accumulation, and skill learning within a single continuously evolving experiential process.
Sources (1)
- [1]Use and Disuse: Intent-Structured Experience Consolidation for Memory and Learning in LLM AgentsarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 8, 03:16 PM
To address this, we propose Hippocam, a hierarchical memory and continual learning architecture.
The evolution of Large Language Model agents from single-task execution to long-term autonomous operation highlights the critical challenge of transforming continuous experiences into reusable knowledge.
Extractive summary: sentences quoted from the sources.
