ZICQ
中 Log in / Sign up
Newsroom Agentic #Agentic Systems #Memory Architecture #System-One/Two #AI Efficiency

Jev-Mem Released: Agentic Memory Architecture Inspired by System-One/Two Cognition

Avatar of Mr.Xu

By Mr.Xu Compiled & Reviewed by Editorial

Published: · 13 views

中文阅读 (Chinese) English Version
Co-reading Select any sentence in the article to highlight, annotate, or copy a quote card. Join discussion ↓

Summary:Jev-Mem is a novel agentic memory architecture inspired by the human cognitive division between System-One (fast decision-making) and System-Two (deliberative reasoning). It introduces a dedicated System-One control plane, a structured multi-relational memory plane, and a System-Two reasoning plane to enhance memory operation efficiency and system performance. On the LoCoMo benchmark, Jev-Mem achieves an LLM-as-a-Judge score of 0.777, an 11.0% improvement over the strongest baseline, while reduc


Key Breakthroughs

Jev-Mem is a novel agentic memory architecture inspired by the human cognitive division between System-One (fast decision-making) and System-Two (deliberative reasoning). The core innovations of this architecture include:

  • System-One Control Plane: Handles rapid decision-making and memory operations, such as memory typing, relational organization, query routing, retrieval-budget allocation, graph traversal, candidate scoring, and adaptive stopping.
  • Multi-Relational Memory Plane: Utilizes a structured storage approach to enhance the accuracy and efficiency of memory retrieval.
  • System-Two Reasoning Plane: Invoked only for complex reasoning and answer synthesis, ensuring system efficiency.

Technical Highlights

  1. Efficiency: The System-One control plane significantly reduces memory construction time and query latency. In the LoCoMo benchmark, memory construction time is reduced to 158 seconds, and average query latency is lowered to 0.93 seconds.
  2. Performance Improvement: Jev-Mem achieves an 11.0% improvement in the LLM-as-a-Judge score over the strongest baseline, reaching 0.777.
  3. Adaptability: The System-Two plane is invoked only for complex tasks, optimizing resource allocation and task processing efficiency.

Industry Impact

The release of Jev-Mem provides AI agents with a more efficient and intelligent solution for long-horizon memory processing and complex task execution. Its design philosophy not only enhances system performance but also opens new possibilities for AI agent applications in resource-constrained environments.

Recommendations for Developers

  • Integration: Developers can integrate Jev-Mem into existing AI agent systems to improve memory operation efficiency and system performance.
  • Optimization Strategies: Leverage the rapid decision-making capabilities of the System-One control plane to optimize the task processing workflow of AI agents.
  • Experimental Validation: Conduct experimental validation in different application scenarios to fully explore the potential of Jev-Mem.

Source: Hugging Face Daily Papers (2026-09-21)

— END —

Tags: #Agentic Systems #Memory Architecture #System-One/Two #AI Efficiency

Editorial & Fact-Checking Note: This article is compiled from primary research, official release documentation, and source papers by the ZICQ Newsroom pipeline with automated entity verification and human editorial review. If you notice any technical inaccuracy, please submit a correction via our corrections policy or email our editorial desk directly.

What's your take on this story?
🚀 Explore & Submit Works

Community Comments

Loading live comments and annotations…