Skip to content
Trending storyDeveloping

Berkeley paper proposes periodic session resets to keep research agents improving

1 article1 sourcesince Oct 15Last article 1h ago ·

Overview

AISummary of 1 article

UC Berkeley researchers propose periodically discarding research agents' session and candidate history and restarting from the best candidate found so far, according to a paper titled Scaling Discovery Requires Forgetting.

The paper says agents that keep their full history become anchored on past attempts and stop improving.

According to a post by Elvis Saravia, the method gives up to 15% over parallel agents and up to 39% over a single agent with the same budget across heuristic programming, kernel optimization, and model training tasks.

Written by AI from the articles below · updated Oct 15, 9:55 PM ET

Check the sources:

Latest developmentOct 15, 9:50 PM ET
UC Berkeley paper finds research agents need periodic forgetting to keep improving

Article timeline

The articles in this story. Times are ET.

Oct 15
  1. elvisX
    UC Berkeley paper finds research agents need periodic forgetting to keep improving

    AIA UC Berkeley paper, Scaling Discovery Requires Forgetting, reports that research agents that keep their full history become anchored on past attempts and stop improving. The authors periodically discard each agent's session and candidate history and restart from the best candidate found so far. The post reports gains of up to 15% over parallel agents and up to 39% over a single agent with the same budget across heuristic programming, kernel optimization, and model training tasks.

    Image from @omarsar0's post

Heat trend

Not enough continuous observations to show a trend yet.