Archive - Reinforced

Positive Gradients, Negative Gradients

+ the Importance of Pre-Training Priors

Dec 19, 2025 • Alex Nikulkov

April 2024

Bandits vs Reinforcement Learning from Human Feedback

Are single-step or multi-step models better suited for RLHF? Find out the main differences between them in this post

Apr 30, 2024 • Alex Nikulkov

Reward Model Overoptimization: Root Causes and Mitigations

When I first ran an RLHF training job, I was surprised at how easily the reward model scores increased during the training process.

Apr 7, 2024 • Alex Nikulkov

January 2024

Reward Modeling for RLHF

An introduction to reward models

Jan 10, 2024 • Alex Nikulkov

Hello World

Welcome to Reinforced!

Jan 7, 2024 • Alex Nikulkov

#nojs-banner { position: fixed; bottom: 0; left: 0; padding: 16px 16px 16px 32px; width: 100%; box-sizing: border-box; background: red; color: white; font-family: -apple-system, "Segoe UI", Roboto, Helvetica, Arial, sans-serif, "Apple Color Emoji", "Segoe UI Emoji", "Segoe UI Symbol"; font-size: 13px; line-height: 13px; } #nojs-banner a { color: inherit; text-decoration: underline; } This site requires JavaScript to run correctly. Please turn on JavaScript or unblock scripts