Model Calibration and Post-Training Effects - Frontier Model Dynamics: Scaling Laws, Calibration, and Post-Training Alignment @ University of Michigan - Ann Arbor

Safety Metrics and Refusal Behavior - Frontier Model Dynamics: Scaling Laws, Calibration, and Post-Training Alignment @ University of Michigan - Ann Arbor

University of Michigan - Ann Arbor

Learners will investigate the foundational dynamics of frontier AI models, focusing on performance extrapolation, calibration shifts, and safety frameworks. You will develop the skills to evaluate scaling laws across compute thresholds and analyze how post-training alignment impacts confidence and capability. Additionally, you will examine critical safety metrics and refusal behaviors to effectively assess the robustness of advanced AI deployments.

Frontier Model Dynamics: Scaling Laws, Calibration, and Post-Training Alignment @ University of Michigan - Ann Arbor

Ch.1 Scaling Dynamics - Frontier Model Dynamics: Scaling Laws, Calibration, and Post-Training Alignment @ University of Michigan - Ann Arbor

pdf material of the course "Frontier Model Dynamics: Scaling Laws, Calibration, and Post-Training Alignment @ University of Michigan - Ann Arbor".

gpt4-selected-pages.pdf

Ch.2 Post-Training Analysis and Safety - Frontier Model Dynamics: Scaling Laws, Calibration, and Post-Training Alignment @ University of Michigan - Ann Arbor

Learn Before

Related

Learn After