Agentic memory speeds up and improves long horizon ai decisions
Jev-Mem: System-One-Controlled Agentic Memory for Efficient AI Agents
Artificial IntelligenceMachine Learning
Summary
Long tasks require AI systems to remember and organize lots of information efficiently. The authors introduce Jev-Mem, a new way for AI agents to manage memory by dividing work between a fast, simple controller and a slower, deeper reasoning process. This method helps the AI quickly decide what memories to keep and retrieve, while only using heavy reasoning when necessary. Their approach made AI memory systems faster and smarter in tests.
What this means in practice
- •For ai system developers: Implement Jev-Mem to build AI agents that manage long-term memory more efficiently, improving both speed and accuracy for complex tasks.
- •For customer service teams: Use AI agents with Jev-Mem memory control to provide faster and more relevant responses in long, multi-turn customer interactions.
Authors
Dongming Jiang, Yi Li, Bingzhe Li
Abstract
Agentic memory is becoming essential for long-horizon AI agents, yet many existing systems rely on autoregressive LLMs to control how memories are organized, retrieved, and used, placing expensive generation on the critical path of memory operations. We introduce \textbf{\method}, a new agentic memory architecture inspired by System-One/System-Two cognition. System One captures fast, lightweight decision-making, whereas System Two performs slower, deliberative reasoning. Jev-Mem brings this division of labor to agentic memory through a dedicated System-One control plane, a structured multi-relational memory plane, and a System-Two reasoning plane. The System-One controller governs memory typing and relational organization during construction, and dynamically performs query routing, retrieval-budget allocation, graph traversal, candidate scoring, and adaptive stopping during retrieval. System Two is invoked only for complex reasoning and answer synthesis. This design improves both memory effectiveness and system efficiency: on LoCoMo Jev-Mem achieves an overall LLM-as-a-Judge score of 0.777, an 11.0\% relative improvement over the strongest baseline, while reducing memory construction time to 158\,s, a 6.6$\times$ speedup over the fastest competing memory system, and lowering average query latency to 0.93\,s, a 36.7\% reduction.