Tutorial Series¶
This series learns through implementation and failure cases. Each entry does more than summarize papers: it places one mechanism back into a shared AGI framework.
Fixed structure¶
- Core question: what was missing from the previous generation?
- Minimal mechanism: run the critical path with as little code as possible.
- Controlled comparison: change one variable at a time.
- Failure case: why did the model fail?
- Current judgment: what does the evidence support?
- Next step: one small, testable question.
Routes¶
- Foundations: Transformer, language modeling, scaling, test-time compute, and post-training.
- Perception: vision–language alignment, multimodal tokens, and long-video understanding.
- Generation: VAE, diffusion, DiT, flow matching, and video world models.
- Action: agents, VLAs, world models, and long-horizon control.
- Closed loop: verifiers, automatic task generation, synthetic data, and self-improving systems.