AI-14 · Research · ENGLISH · PLANNED
Begin with a baseline, trace a model, measure errors and improve it.
Study time is an initial activity estimate, excluding extra reading, revision and extended laboratory or project work. Videos and question bodies have not been produced by adding this outline.
Complete these courses or demonstrate equivalent knowledge. These are recommended prerequisites, not enforced enrolment locks.
YouTube playlist: ARK Artificial Intelligence | Advanced AI and Research Studio
Playlist title reserved in the plan; no YouTube playlist or upload has been created yet.
Sign in to save this learning pathPractical outcome: Produce a prerequisite map for one advanced learning paper.
Explain statistical learning theory using a visual and one worked real-world example. Context: Produce a prerequisite map for one advanced learning paper. End with a prediction or check-for-understanding.
Explain generalisation bounds using a visual and one worked real-world example. Context: Produce a prerequisite map for one advanced learning paper. End with a prediction or check-for-understanding.
Explain kernel methods using a visual and one worked real-world example. Context: Produce a prerequisite map for one advanced learning paper. End with a prediction or check-for-understanding.
Explain optimal transport entry using a visual and one worked real-world example. Context: Produce a prerequisite map for one advanced learning paper. End with a prediction or check-for-understanding.
Use the first concepts from this section. Produce a prerequisite map for one advanced learning paper. Produce a labelled diagram, trace or calculation; explain each step and check one changed input.
3–6 questions on the preceding first-half concepts only: recall, visual interpretation, application and one plausible misconception. Explain each answer after submission.
Explain information bottleneck entry using a visual and one worked real-world example. Context: Produce a prerequisite map for one advanced learning paper. End with a prediction or check-for-understanding.
Explain graph neural networks using a visual and one worked real-world example. Context: Produce a prerequisite map for one advanced learning paper. End with a prediction or check-for-understanding.
Explain neuro-symbolic systems using a visual and one worked real-world example. Context: Produce a prerequisite map for one advanced learning paper. End with a prediction or check-for-understanding.
Explain mechanistic interpretability using a visual and one worked real-world example. Context: Produce a prerequisite map for one advanced learning paper. End with a prediction or check-for-understanding.
Produce a prerequisite map for one advanced learning paper. Work without the example, compare two cases, record assumptions and limitations, then self-review correctness, reasoning and clarity.
6–12 questions covering Statistical learning theory, Generalisation bounds, Kernel methods and the remaining section concepts, with cumulative retrieval from earlier sections. Include a new transfer problem; 70% unlocks progression.
Practical outcome: Compare evidence for two proposed approaches to one open problem.
Explain self-supervised objectives using a visual and one worked real-world example. Context: Compare evidence for two proposed approaches to one open problem. End with a prediction or check-for-understanding.
Explain world-model concepts using a visual and one worked real-world example. Context: Compare evidence for two proposed approaches to one open problem. End with a prediction or check-for-understanding.
Explain continual learning using a visual and one worked real-world example. Context: Compare evidence for two proposed approaches to one open problem. End with a prediction or check-for-understanding.
Explain federated learning using a visual and one worked real-world example. Context: Compare evidence for two proposed approaches to one open problem. End with a prediction or check-for-understanding.
Use the first concepts from this section. Compare evidence for two proposed approaches to one open problem. Produce a labelled diagram, trace or calculation; explain each step and check one changed input.
3–6 questions on the preceding first-half concepts only: recall, visual interpretation, application and one plausible misconception. Explain each answer after submission.
Explain scientific machine learning using a visual and one worked real-world example. Context: Compare evidence for two proposed approaches to one open problem. End with a prediction or check-for-understanding.
Explain embodied intelligence using a visual and one worked real-world example. Context: Compare evidence for two proposed approaches to one open problem. End with a prediction or check-for-understanding.
Explain alignment evaluation using a visual and one worked real-world example. Context: Compare evidence for two proposed approaches to one open problem. End with a prediction or check-for-understanding.
Explain capability claims versus evidence using a visual and one worked real-world example. Context: Compare evidence for two proposed approaches to one open problem. End with a prediction or check-for-understanding.
Compare evidence for two proposed approaches to one open problem. Work without the example, compare two cases, record assumptions and limitations, then self-review correctness, reasoning and clarity.
6–12 questions covering Self-supervised objectives, World-model concepts, Continual learning and the remaining section concepts, with cumulative retrieval from earlier sections. Include a new transfer problem; 70% unlocks progression.
Practical outcome: Reproduce a modest result on a bounded public or synthetic dataset.
Explain literature search using a visual and one worked real-world example. Context: Reproduce a modest result on a bounded public or synthetic dataset. End with a prediction or check-for-understanding.
Explain hypothesis specification using a visual and one worked real-world example. Context: Reproduce a modest result on a bounded public or synthetic dataset. End with a prediction or check-for-understanding.
Explain baseline replication using a visual and one worked real-world example. Context: Reproduce a modest result on a bounded public or synthetic dataset. End with a prediction or check-for-understanding.
Explain ablation design using a visual and one worked real-world example. Context: Reproduce a modest result on a bounded public or synthetic dataset. End with a prediction or check-for-understanding.
Use the first concepts from this section. Reproduce a modest result on a bounded public or synthetic dataset. Produce a labelled diagram, trace or calculation; explain each step and check one changed input.
3–6 questions on the preceding first-half concepts only: recall, visual interpretation, application and one plausible misconception. Explain each answer after submission.
Explain compute planning in usd using a visual and one worked real-world example. Context: Reproduce a modest result on a bounded public or synthetic dataset. End with a prediction or check-for-understanding.
Explain reproducible artifacts using a visual and one worked real-world example. Context: Reproduce a modest result on a bounded public or synthetic dataset. End with a prediction or check-for-understanding.
Explain failure-case analysis using a visual and one worked real-world example. Context: Reproduce a modest result on a bounded public or synthetic dataset. End with a prediction or check-for-understanding.
Explain research writing using a visual and one worked real-world example. Context: Reproduce a modest result on a bounded public or synthetic dataset. End with a prediction or check-for-understanding.
Explain peer review using a visual and one worked real-world example. Context: Reproduce a modest result on a bounded public or synthetic dataset. End with a prediction or check-for-understanding.
Reproduce a modest result on a bounded public or synthetic dataset. Work without the example, compare two cases, record assumptions and limitations, then self-review correctness, reasoning and clarity.
6–12 questions covering Literature search, Hypothesis specification, Baseline replication and the remaining section concepts, with cumulative retrieval from earlier sections. Include a new transfer problem; 70% unlocks progression.
Combine all three sections: Reproduce a modest result on a bounded public or synthetic dataset. Submit an original worked artifact, evidence or tests, assumptions and a reflection. Rubric: correctness, method, practical usefulness, reproducibility and communication. Self-reviewed, not certified.
6–12 balanced questions sampling every section of Advanced AI and Research Studio, including interpretation, calculation or reasoning, a misconception and a new application. Passing is 70%; review errors before retrying.
Data experiments, evaluations, model tracing and reproducible projects. Guided work leads to an independent attempt, with a diagram, explanation, trace, calculation or tested artifact. Question bodies are generated when opened in your private learning path. Programming submissions are stored as text; the platform does not execute submitted code.
Use original or appropriately licensed datasets. Practical science activities use models or supplied data unless a qualified supervisor has reviewed a physical experiment.
Open the interactive practice sampler →