Stage 14 · Live
Advanced Robotics / Reinforcement Learning
Parts 1–2: RL foundations for robots (MDP, policy/value, exploration, reward shaping) plus value-based deep RL / DQN intuition (why tables break, function approximation, Q-net + replay + target, stability traps). Soft-recommend Stages 11–13; hard locks cleared; no payment gating. Browser educational toys only — NOT MuJoCo/Isaac/stable-baselines. Parts 3–4 (policy gradients, sim-to-real caveats) planned later.
Journey · Stage 13 Perception/Planning → Stage 14 Advanced Robotics / Reinforcement Learning (Parts 1–2: foundations → DQN intuition / value-based deep RL) → Parts 3–4 planned (policy gradients / sim-to-real). Soft-recommend Stages 11–13. Bridges classical stack (Stages 7–13) into learning intuition. Open now. Interactives are educational browser RL toys — NOT MuJoCo / Isaac / stable-baselines.
Lessons
- 01
Why Robots Learn
60–75 min
- 02
States, Actions & Rewards
60–75 min
- 03
Policies & Value
60–75 min
- 04
Episodes, Exploration & Exploitation
70–85 min
- 05
Reward Shaping for Manipulation
70–85 min
- 06
Why Tabular Q Breaks
60–75 min
- 07
Function Approximation Intuition
65–80 min
- 08
DQN Core Ideas
70–85 min
- 09
Stability Traps & Overestimation
65–80 min
- 10
DQN Intuition Challenge
75–95 min