In this episode, we unpack why sparse, final-only rewards hobble reinforcement learning in large language models and how dense rewards via a process reward model act like a patient teacher, giving praise for micro-steps along the way. We explore how fortifying these steps reshapes the model’s reasoning, why broad, inconsistent feedback can cause global unlearning, and what this means for building AI that can truly reason across domains.
Note: This podcast was AI-generated, and sometimes AI can make mistakes. Please double-check any critical information.
Podden och tillhörande omslagsbild på den här sidan tillhör
Mike Breault. Innehållet i podden är skapat av Mike Breault och inte av,
eller tillsammans med, Poddtoppen.
