ResearcharXivNEW

On the Fragility of Self-Improving Agents: Variance, Task Order, and Underspecification

Ye 2026-08-18
Qinyuan YeYu LiYada Pruksachatkun

Memory-based self-improving agents--those that learn from an online stream of tasks and improve over time by maintaining a textual memory bank--have shown great promise in recent literature. However, the reliability aspects of these methods have been critically overlooked. In this work, we conduct a comprehensive re-evaluation of two memory-based methods, broadening the scope of evaluation along t

Read on arXiv
Data aggregated and editorially reviewed by TrendMing.

Key Contributions

  • Memory-based self-improving agents--those that learn from an online stream of tasks and improve over time by maintaining a textual memory bank--have shown great promise in recent literature.
  • However, the reliability aspects of these methods have been critically overlooked.
  • In this work, we conduct a comprehensive re-evaluation of two memory-based methods, broadening the scope of evaluation along t

Research Themes

AIResearch