Subjects
Pick a skill, or let the selector choose — it aims at questions you have roughly a 78% chance of getting right.
AGI Agents (6-week course)
The architecture of general-purpose agents — world models, memory, RL, ReAct, multi-agent teams, alignment and honest evaluation. Mirrors your AGI-Curriculum repo week by week.
Practise
W0 · Setup, tests & model backends
Not started
5 questions
W1 · World models & planning
Not started
5 questions
W1 · Memory systems
Not started
5 questions
W1 · Cognitive architectures
Not started
5 questions
W2 · MDPs & value functions
Not started
5 questions
W2 · Q-learning & SARSA
Not started
5 questions
W2 · Dyna, MCTS & the DeepMind lineage
Not started
5 questions
W3 · ReAct & tool use
Not started
5 questions
W3 · Stateful graphs, memory & planning
Not started
5 questions
W3 · Traces, failures & self-improvement
Not started
5 questions
W4 · Multi-agent coordination
Not started
5 questions
W4 · Negotiation, voting & debate
Not started
5 questions
W4 · AutoGen & CrewAI in practice
Not started
5 questions
W5 · Constitutional AI
Not started
5 questions
W5 · Reward models, DPO & reward hacking
Not started
5 questions
W5 · Guardrails, injection & red-teaming
Not started
5 questions
W6 · Integrating the agent
Not started
5 questions
W6 · Evaluating agents honestly
Not started
5 questions
W6 · Honest scope & what AGI means
Not started
5 questions
Probability & Statistics
Describing data, reasoning about chance, and the difference between a sample and the world.
Practise
JavaScript
Types and coercion, arrays, closures, asynchrony, and the complexity of what you just wrote.
Practise