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

Add to Chrome

Add to Firefox

Picture for Siddhartha Brahma

Siddhartha Brahma

Breaking BERT: Evaluating and Optimizing Sparsified Attention

Oct 07, 2022
Siddhartha Brahma, Polina Zablotskaia, David Mimno

Figure 1 for Breaking BERT: Evaluating and Optimizing Sparsified Attention

Figure 2 for Breaking BERT: Evaluating and Optimizing Sparsified Attention

Figure 3 for Breaking BERT: Evaluating and Optimizing Sparsified Attention

Figure 4 for Breaking BERT: Evaluating and Optimizing Sparsified Attention

* Shorter version accepted to SNN2021 workshop

Access Paper or Ask Questions

CLAR: A Cross-Lingual Argument Regularizer for Semantic Role Labeling

Nov 09, 2020
Ishan Jindal, Yunyao Li, Siddhartha Brahma, Huaiyu Zhu

Figure 1 for CLAR: A Cross-Lingual Argument Regularizer for Semantic Role Labeling

Figure 2 for CLAR: A Cross-Lingual Argument Regularizer for Semantic Role Labeling

Figure 3 for CLAR: A Cross-Lingual Argument Regularizer for Semantic Role Labeling

Figure 4 for CLAR: A Cross-Lingual Argument Regularizer for Semantic Role Labeling

* EMNLP 2020, ACL Findings

Access Paper or Ask Questions

Small but Mighty: New Benchmarks for Split and Rephrase

Sep 17, 2020
Li Zhang, Huaiyu Zhu, Siddhartha Brahma, Yunyao Li

Figure 1 for Small but Mighty: New Benchmarks for Split and Rephrase

Figure 2 for Small but Mighty: New Benchmarks for Split and Rephrase

Figure 3 for Small but Mighty: New Benchmarks for Split and Rephrase

Figure 4 for Small but Mighty: New Benchmarks for Split and Rephrase

* In EMNLP 2020

Access Paper or Ask Questions

Improved Language Modeling by Decoding the Past

Sep 29, 2018
Siddhartha Brahma

Figure 1 for Improved Language Modeling by Decoding the Past

Figure 2 for Improved Language Modeling by Decoding the Past

Access Paper or Ask Questions

Unsupervised Learning of Sentence Representations Using Sequence Consistency

Sep 29, 2018
Siddhartha Brahma

Figure 1 for Unsupervised Learning of Sentence Representations Using Sequence Consistency

Figure 2 for Unsupervised Learning of Sentence Representations Using Sequence Consistency

Figure 3 for Unsupervised Learning of Sentence Representations Using Sequence Consistency

Figure 4 for Unsupervised Learning of Sentence Representations Using Sequence Consistency

* Submitted to ICLR 2019

Access Paper or Ask Questions

Improved Sentence Modeling using Suffix Bidirectional LSTM

Sep 10, 2018
Siddhartha Brahma

Figure 1 for Improved Sentence Modeling using Suffix Bidirectional LSTM

Figure 2 for Improved Sentence Modeling using Suffix Bidirectional LSTM

Figure 3 for Improved Sentence Modeling using Suffix Bidirectional LSTM

Figure 4 for Improved Sentence Modeling using Suffix Bidirectional LSTM

Access Paper or Ask Questions