ZerotoMastery – AI for Beginners: Reasoning Models

Understand how Reasoning Models actually work from scratch with practical and fun lessons. You’ll go behind the curtain on one of the biggest breakthroughs in AI and learn how reasoning models “think”, how they’re trained with reinforcement learning and dive deep into datasets like PRM800K, and why reasoning models may be full of secrets and lies…

What you’ll learn
What “reasoning models” really are (and aren’t)
Compare reasoning styles across models
The PRM800K dataset that powered this breakthrough
Prompt and context engineering built for reasoning models
The importance of the Generator–Verifier Gap (GVG)
How Reinforcement Learning + Process Reward Models shape reasoning
Dive deep into AI research papers
Why reasoning models may be lying to you…
Why Learn About Reasoning Models?
Reasoning models are one of the biggest breakthroughs in AI as they give models the ability to think through things using a “scratchpad” of sorts, introducing something more akin to System 2 thinking.

Reasoning models feel like magic…until they don’t. This course helps you understand what’s happening under the hood, so you know how to use them, why they work, and where this technology is going.

You’ll start with the basics, then quickly move into real model behavior. You’ll practice your skills and explore reasoning models with real, hands-on exercises and practice.

Then you’ll connect the dots to how these systems are trained: reinforcement learning, RLHF, process reward models, the PRM800K dataset, and the impact of the latest scaling law on the future of AI: test-time compute.

You’ll also explore the tricky (but maybe more interesting)part: when reasoning models lie to you and the secrets they keep.

https://anonymz.com/?https://zerotomastery.io/courses/llm-reasoning-models/