VFlash: Fast, Accurate Visual Planning via Flash Primitive Decoding
u_Unrolled
All paths

Path · Advanced

Reinforcement Learning

From bandits and Bellman to PPO, then GRPO and verifiable rewards: how frontier labs train reasoning models.

7 lessons3 stagesabout 4 hours3 free
Start with “Probability & Statistics”
You will build: Teach a model maths with GRPO

The roadmap

  1. 1

    Probability first

  2. 2

    Reinforcement learning

  3. 3

    RL for language models

Every lesson in this path is written.

Practise what you learned on the AI Engineer Sheet, or build one of the projects.

Explore other paths