Representation learning notes for static word vectors and embedding objectives.
Notes
| Area | Note | Status |
|---|---|---|
| Overview | Word Embeddings | draft |
| Static embedding | Word2Vec | draft |
| Optimization | Negative Sampling | draft |
| Static embedding | GloVe | draft |
Learning Use
- Understand how discrete tokens become continuous vectors.
- Connect local prediction objectives with global co-occurrence statistics.
- Prepare for contextual embeddings in pretrained language models.