You have to present one paper on
January 8, 2020 January 15, 2020.
The presentation should last 15-20 minutes.
Some paper might be a little bit difficult, it is ok to not fully understand each equation as long as you are able to explain intuitions.
You can present the paper either on black board or using slides.
There should be 3 or 4 students per group. Groups can be different from lab exercises and project groups. Group and papers should be choose before December 7, 2019. You have to send me one mail per group where all group members are in the mail recipients. Papers will be distributed in a first come first served basis.
- Recurrent Highway Networks (Zilly et al.) [PDF]
- Fast and Accurate Entity Recognition with Iterated Dilated Convolutions (Strubell et al.) [PDF]
- Learning with Noise-Contrastive Estimation: Easing training by learning to scale (Labeau and Allauzen) [PDF]
- Analyzing Multi-Head Self-Attention: Specialized Heads Do the Heavy Lifting, the Rest Can Be Pruned (Voita et al.) [PDF]
- Adaptively Sparse Transformers (Correia et al.) [PDF]
Improving Neural Machine Translation Models with Monolingual Data (Sennrich et al.) [PDF]
Graph Convolutional Encodersfor Syntax-aware Neural Machine Translation (Bastings et al.) [PDF]
On the Limitations of Unsupervised Bilingual Dictionary Induction (Søgaard et al.) [PDF]
Deep Pyramid Convolutional Neural Networks for Text Categorization (Johnson and Zhang) [PDF]
Attention is not not Explanation (Wiegreffe and Pinter) [PDF]
Step-by-Step: Separating Planning from Realization in Neural Data-to-Text Generation (Moryossef et al.) [PDF]