Syllabus · Post Training

A rigorous, hands-on exploration of post-training — how we shape model behavior through reinforcement learning, align objectives, design reward functions, build evaluations, and turn foundation models into reliable, useful AI systems. The hard problem in AI isn't making machines smarter, it's teaching them to handle problems without right answers. Understanding the full training pipeline from pretraining to deployment. No class this week. Enjoy your break! Present your final projects to the class.

saved by

• Dante Maggiotto