Algorithmic Model Alignment: The Math Behind Safety Parameters
A mathematical investigation into the safety parameters of large language models, explaining the mechanics of RLHF and DPO.
Showing 1–2 of 2 articles
A mathematical investigation into the safety parameters of large language models, explaining the mechanics of RLHF and DPO.
A mathematical guide to the mechanics of loss functions, showing how algorithms measure optimization errors and adjust weights.