Test-Time Reasoning for LLMs

Learn why giving language models more inference-time computation can improve reasoning, how chain-of-thought methods use that budget, and where the approach breaks down. This course is for learners who know the basics of LLMs and want a practical map of modern reasoning techniques.

4 sections ยท 11 lessons

Course outline

Why It Helps

  1. Slow Thinking for Models
  2. Compute as a Budget
  3. Thoughts as Hidden Variables

Token Reasoning

  1. Chain-of-Thought Foundations
  2. Branching and Revision
  3. Reinforcement for Reasoning

Trust and Tools

  1. Tool-Augmented Reasoning
  2. Faithfulness Under Pressure

Beyond Tokens

  1. Continuous Thinking Mechanisms
  2. Latent Training Loops
  3. Scaling Thinking Time

Start learning with Wondering