AI-08 · Advanced · 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 | Natural Language Processing and Transformers
Playlist title reserved in the plan; no YouTube playlist or upload has been created yet.
Sign in to save this learning pathPractical outcome: Compare representations of a short original text corpus.
Explain tokenisation using a visual and one worked real-world example. Context: Compare representations of a short original text corpus. End with a prediction or check-for-understanding.
Explain subword units using a visual and one worked real-world example. Context: Compare representations of a short original text corpus. End with a prediction or check-for-understanding.
Explain bag of words using a visual and one worked real-world example. Context: Compare representations of a short original text corpus. End with a prediction or check-for-understanding.
Explain tf-idf using a visual and one worked real-world example. Context: Compare representations of a short original text corpus. End with a prediction or check-for-understanding.
Use the first concepts from this section. Compare representations of a short original text corpus. 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 word embeddings using a visual and one worked real-world example. Context: Compare representations of a short original text corpus. End with a prediction or check-for-understanding.
Explain sequence models using a visual and one worked real-world example. Context: Compare representations of a short original text corpus. End with a prediction or check-for-understanding.
Explain recurrent networks using a visual and one worked real-world example. Context: Compare representations of a short original text corpus. End with a prediction or check-for-understanding.
Explain language modelling using a visual and one worked real-world example. Context: Compare representations of a short original text corpus. End with a prediction or check-for-understanding.
Compare representations of a short original text corpus. Work without the example, compare two cases, record assumptions and limitations, then self-review correctness, reasoning and clarity.
6–12 questions covering Tokenisation, Subword units, Bag of words and the remaining section concepts, with cumulative retrieval from earlier sections. Include a new transfer problem; 70% unlocks progression.
Practical outcome: Trace attention weights in a tiny sentence example.
Explain queries keys and values using a visual and one worked real-world example. Context: Trace attention weights in a tiny sentence example. End with a prediction or check-for-understanding.
Explain scaled dot-product attention using a visual and one worked real-world example. Context: Trace attention weights in a tiny sentence example. End with a prediction or check-for-understanding.
Explain attention masks using a visual and one worked real-world example. Context: Trace attention weights in a tiny sentence example. End with a prediction or check-for-understanding.
Explain multi-head attention using a visual and one worked real-world example. Context: Trace attention weights in a tiny sentence example. End with a prediction or check-for-understanding.
Use the first concepts from this section. Trace attention weights in a tiny sentence example. 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 positional information using a visual and one worked real-world example. Context: Trace attention weights in a tiny sentence example. End with a prediction or check-for-understanding.
Explain residual connections using a visual and one worked real-world example. Context: Trace attention weights in a tiny sentence example. End with a prediction or check-for-understanding.
Explain transformer blocks using a visual and one worked real-world example. Context: Trace attention weights in a tiny sentence example. End with a prediction or check-for-understanding.
Explain autoregressive decoding using a visual and one worked real-world example. Context: Trace attention weights in a tiny sentence example. End with a prediction or check-for-understanding.
Trace attention weights in a tiny sentence example. Work without the example, compare two cases, record assumptions and limitations, then self-review correctness, reasoning and clarity.
6–12 questions covering Queries keys and values, Scaled dot-product attention, Attention masks and the remaining section concepts, with cumulative retrieval from earlier sections. Include a new transfer problem; 70% unlocks progression.
Practical outcome: Evaluate a classifier and a generator using separate error criteria.
Explain text classification using a visual and one worked real-world example. Context: Evaluate a classifier and a generator using separate error criteria. End with a prediction or check-for-understanding.
Explain named entities using a visual and one worked real-world example. Context: Evaluate a classifier and a generator using separate error criteria. End with a prediction or check-for-understanding.
Explain information extraction using a visual and one worked real-world example. Context: Evaluate a classifier and a generator using separate error criteria. End with a prediction or check-for-understanding.
Explain translation evaluation using a visual and one worked real-world example. Context: Evaluate a classifier and a generator using separate error criteria. End with a prediction or check-for-understanding.
Use the first concepts from this section. Evaluate a classifier and a generator using separate error criteria. 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 summarisation evaluation using a visual and one worked real-world example. Context: Evaluate a classifier and a generator using separate error criteria. End with a prediction or check-for-understanding.
Explain perplexity using a visual and one worked real-world example. Context: Evaluate a classifier and a generator using separate error criteria. End with a prediction or check-for-understanding.
Explain factuality checks using a visual and one worked real-world example. Context: Evaluate a classifier and a generator using separate error criteria. End with a prediction or check-for-understanding.
Explain dataset contamination using a visual and one worked real-world example. Context: Evaluate a classifier and a generator using separate error criteria. End with a prediction or check-for-understanding.
Evaluate a classifier and a generator using separate error criteria. Work without the example, compare two cases, record assumptions and limitations, then self-review correctness, reasoning and clarity.
6–12 questions covering Text classification, Named entities, Information extraction and the remaining section concepts, with cumulative retrieval from earlier sections. Include a new transfer problem; 70% unlocks progression.
Combine all three sections: Evaluate a classifier and a generator using separate error criteria. 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 Natural Language Processing and Transformers, 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 →