Get our free extension to see links to code for papers anywhere online!

Add to Chrome

Add to Firefox

Picture for Mrinmaya Sachan

Mrinmaya Sachan

Differentiable Subset Pruning of Transformer Heads

Aug 22, 2021
Jiaoda Li, Ryan Cotterell, Mrinmaya Sachan

Figure 1 for Differentiable Subset Pruning of Transformer Heads

Figure 2 for Differentiable Subset Pruning of Transformer Heads

Figure 3 for Differentiable Subset Pruning of Transformer Heads

Figure 4 for Differentiable Subset Pruning of Transformer Heads

* TACL 2021

Access Paper or Ask Questions

How Good Is NLP? A Sober Look at NLP Tasks through the Lens of Social Impact

Jun 04, 2021
Zhijing Jin, Geeticka Chauhan, Brian Tse, Mrinmaya Sachan, Rada Mihalcea

Figure 1 for How Good Is NLP? A Sober Look at NLP Tasks through the Lens of Social Impact

Figure 2 for How Good Is NLP? A Sober Look at NLP Tasks through the Lens of Social Impact

Figure 3 for How Good Is NLP? A Sober Look at NLP Tasks through the Lens of Social Impact

Figure 4 for How Good Is NLP? A Sober Look at NLP Tasks through the Lens of Social Impact

* ACL 2021 Findings

Access Paper or Ask Questions

Bird's Eye: Probing for Linguistic Graph Structures with a Simple Information-Theoretic Approach

May 25, 2021
Yifan Hou, Mrinmaya Sachan

Figure 1 for Bird's Eye: Probing for Linguistic Graph Structures with a Simple Information-Theoretic Approach

Figure 2 for Bird's Eye: Probing for Linguistic Graph Structures with a Simple Information-Theoretic Approach

Figure 3 for Bird's Eye: Probing for Linguistic Graph Structures with a Simple Information-Theoretic Approach

Figure 4 for Bird's Eye: Probing for Linguistic Graph Structures with a Simple Information-Theoretic Approach

* Accepted for publication at ACL 2021. This is the camera ready version. Our implementation is available in https://github.com/yifan-h/Graph_Probe-Birds_Eye

Access Paper or Ask Questions

Stronger Transformers for Neural Multi-Hop Question Generation

Oct 22, 2020
Devendra Singh Sachan, Lingfei Wu, Mrinmaya Sachan, William Hamilton

Figure 1 for Stronger Transformers for Neural Multi-Hop Question Generation

Figure 2 for Stronger Transformers for Neural Multi-Hop Question Generation

Figure 3 for Stronger Transformers for Neural Multi-Hop Question Generation

Figure 4 for Stronger Transformers for Neural Multi-Hop Question Generation

* Code will be made available

Access Paper or Ask Questions