Case Study

Diagnosing Unstable Reinforcement Learning Training

Given the following training scenario, identify the most probable underlying issue with the gradient estimation process and explain why it leads to the observed unstable performance.

0

1

Updated 2025-10-02

Contributors are:

Who are from:

Tags

Ch.4 Alignment - Foundations of Large Language Models

Foundations of Large Language Models

Foundations of Large Language Models Course

Computing Sciences

Analysis in Bloom's Taxonomy

Cognitive Psychology

Psychology

Social Science

Empirical Science

Science