Development

A Meta-RL framework lets agents learn from feedback without gradient updates

development2025-12-22

Claim

Swiss researchers unveiled a Meta-RL framework letting agents reflect on feedback without gradient updates, giving double-digit gains by turning test-time compute into active exploration.

Reporting date: 2025-12-22 (not necessarily the event date)

Sources

Attribution and reporting linked by this record; links do not imply independent verification.

Welcome to December 22, 2025 (theinnermostloop.substack.com)

Knowledge graphopen in browser →

Reported in

Welcome to December 22, 2025

Evidence for

Scaffolding, not weights

Referenced by

Welcome to December 22, 2025

Cite or correct this record

Citation formatCitation text

A Meta-RL framework lets agents learn from feedback without gradient updates
Innermost Loop KB — secondary record
Reporting date: 2025-12-22
KB permalink: https://nicholsn.github.io/innermost-loop-kb/developments/2025-12-22-meta-rl-without-gradients
Claim: Swiss researchers unveiled a Meta-RL framework letting agents reflect on feedback without gradient updates, giving double-digit gains by turning test-time compute into active exploration.
Attributed sources:
- Welcome to December 22, 2025 — https://theinnermostloop.substack.com/p/welcome-to-december-22-2025
This citation does not imply independent verification of linked-source claims.

Suggest a correction — opens a GitHub draft for you to review and submit.