Week 08 · overview
Week 8: Language Modeling and Recurrent Neural Networks
Language models learn patterns in token sequences. Recurrent neural networks process those sequences in order, carrying a state forward so later steps can depend on earlier input.
| Day | Role | Evidence |
|---|---|---|
| 1 | Lesson 1: Language modeling | Distinguish n-gram, CBOW, and skip-gram objectives |
| 2 | Lesson 2: Recurrent networks | Trace input and state through a shared recurrent cell |
| 3 | Investigation | Calculate state updates and test order dependence |
| 4 | Lab | Compare deterministic recurrent-state traces |
| 5 | Assessment + Reflection | Repair a state-means-understanding claim |
Your Recurrent State Trace Record preserves each token value, prior state, updated state, candidate score, comparison, claim repair, and limitation. The Lab is a deterministic teaching simulation, not a trained language model.
The complete pinned Microsoft source and credits remain in upstream/.
Review the source record.
Chapter question
How does Language Modeling and Recurrent Neural Networks change what evidence a student should trust?
What you will understand
You will connect the named mechanism to the evidence it can and cannot support.
What you will do
You will inspect supplied cases, trace the important calculation, and test one bounded claim.
What you will produce
You will produce the week evidence record named in the Lab or reflection.
How the week connects
This week carries the previous AI mechanism forward into a stricter evidence check.
Key vocabulary
mechanism; evidence; boundary; test; claim
Approximate time
Five class meetings at about 210 instructional minutes total.
Final evidence required
Submit the completed week record with at least two exact labels from the artifact.