What changed

Changes since HEAD · 3 pages changed

Legend: added · removed. Page titles link to the full preview.

Paper changed

LaTeX edits are highlighted; other paper inputs may not be.

Paper
Open the paper PDF

Slides · Asta workspace example – Attention in sequence models changed

1 changed or new; 0 removed; 10 slides
Expand 4 unchanged slides (1–4)

Slide 5: Attention weights the available context — changed

Changes marked
Clean
Before
After
Expand 5 unchanged slides (6–10)

A small literature review – Asta workspace example changed

lit-review/index.html

A small literature review

This literature review is the editable Quarto source. To get a PDF or LaTeX source, render it with Quarto (see the workspace guide). This review cites the project-root references.bib. The separate paper/main.tex is edited directly and paired with Overleaf, so it keeps its own copy of the entries it cites.

Example claimclaims

Vaswani and colleagues proposed a Transformer architecture based solely on attention mechanisms“We propose a new simple network architecture, the Transformer, based solely on attention mechanisms”— (Vaswani et al. 2017, abstract) (Vaswani et al. 2017). TheBERT later applied the Transformer encoder to bidirectional language representationevidence added“We introduce a new language representation model called BERT, which stands for Bidirectional Encoder Representations from Transformers.”— (Devlin et al. 2019, abstract) (Devlin et al. 2019). Each highlighted claim exposesshows its source quote on hover or keyboard focus.

References

Devlin, Jacob, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. “BERT: Pre-Training of Deep Bidirectional Transformers for Language Understanding.” https://arxiv.org/abs/1810.04805.
Vaswani, Ashish, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Lukasz Kaiser, and Illia Polosukhin. 2017. “Attention Is All You Need.” https://arxiv.org/abs/1706.03762.