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

Chrome logo Add to Chrome

Firefox logo Add to Firefox
Picture for Nitish Shirish Keskar

Nitish Shirish Keskar

Modeling Multi-hop Question Answering as Single Sequence Prediction


May 18, 2022
Semih Yavuz, Kazuma Hashimoto, Yingbo Zhou, Nitish Shirish Keskar, Caiming Xiong

Figure 1 for Modeling Multi-hop Question Answering as Single Sequence Prediction
Figure 2 for Modeling Multi-hop Question Answering as Single Sequence Prediction
Figure 3 for Modeling Multi-hop Question Answering as Single Sequence Prediction
Figure 4 for Modeling Multi-hop Question Answering as Single Sequence Prediction
* ACL 2022 
  Access Paper or Ask Questions

Combining Data-driven Supervision with Human-in-the-loop Feedback for Entity Resolution


Nov 20, 2021
Wenpeng Yin, Shelby Heinecke, Jia Li, Nitish Shirish Keskar, Michael Jones, Shouzhong Shi, Stanislav Georgiev, Kurt Milich, Joseph Esposito, Caiming Xiong

Figure 1 for Combining Data-driven Supervision with Human-in-the-loop Feedback for Entity Resolution
Figure 2 for Combining Data-driven Supervision with Human-in-the-loop Feedback for Entity Resolution
* Camera-ready for Data-Centric AI Workshop at NeurIPS 2021 
  Access Paper or Ask Questions

Unsupervised Paraphrase Generation via Dynamic Blocking


Oct 24, 2020
Tong Niu, Semih Yavuz, Yingbo Zhou, Huan Wang, Nitish Shirish Keskar, Caiming Xiong

Figure 1 for Unsupervised Paraphrase Generation via Dynamic Blocking
Figure 2 for Unsupervised Paraphrase Generation via Dynamic Blocking
Figure 3 for Unsupervised Paraphrase Generation via Dynamic Blocking
Figure 4 for Unsupervised Paraphrase Generation via Dynamic Blocking
* 10 pages 
  Access Paper or Ask Questions

GeDi: Generative Discriminator Guided Sequence Generation


Sep 14, 2020
Ben Krause, Akhilesh Deepak Gotmare, Bryan McCann, Nitish Shirish Keskar, Shafiq Joty, Richard Socher, Nazneen Fatema Rajani

Figure 1 for GeDi: Generative Discriminator Guided Sequence Generation
Figure 2 for GeDi: Generative Discriminator Guided Sequence Generation
Figure 3 for GeDi: Generative Discriminator Guided Sequence Generation
Figure 4 for GeDi: Generative Discriminator Guided Sequence Generation

  Access Paper or Ask Questions

Mirostat: A Perplexity-Controlled Neural Text Decoding Algorithm


Jul 29, 2020
Sourya Basu, Govardana Sachitanandam Ramachandran, Nitish Shirish Keskar, Lav R. Varshney

Figure 1 for Mirostat: A Perplexity-Controlled Neural Text Decoding Algorithm
Figure 2 for Mirostat: A Perplexity-Controlled Neural Text Decoding Algorithm
Figure 3 for Mirostat: A Perplexity-Controlled Neural Text Decoding Algorithm
Figure 4 for Mirostat: A Perplexity-Controlled Neural Text Decoding Algorithm
* 18 pages, 8 figures 
  Access Paper or Ask Questions

ProGen: Language Modeling for Protein Generation


Mar 08, 2020
Ali Madani, Bryan McCann, Nikhil Naik, Nitish Shirish Keskar, Namrata Anand, Raphael R. Eguchi, Po-Ssu Huang, Richard Socher


  Access Paper or Ask Questions

Limits of Detecting Text Generated by Large-Scale Language Models


Feb 09, 2020
Lav R. Varshney, Nitish Shirish Keskar, Richard Socher

* ITA 2020 
  Access Paper or Ask Questions

Global Capacity Measures for Deep ReLU Networks via Path Sampling


Oct 22, 2019
Ryan Theisen, Jason M. Klusowski, Huan Wang, Nitish Shirish Keskar, Caiming Xiong, Richard Socher

Figure 1 for Global Capacity Measures for Deep ReLU Networks via Path Sampling

  Access Paper or Ask Questions

CTRL: A Conditional Transformer Language Model for Controllable Generation


Sep 20, 2019
Nitish Shirish Keskar, Bryan McCann, Lav R. Varshney, Caiming Xiong, Richard Socher

Figure 1 for CTRL: A Conditional Transformer Language Model for Controllable Generation
Figure 2 for CTRL: A Conditional Transformer Language Model for Controllable Generation
Figure 3 for CTRL: A Conditional Transformer Language Model for Controllable Generation
Figure 4 for CTRL: A Conditional Transformer Language Model for Controllable Generation

  Access Paper or Ask Questions

Pretrained AI Models: Performativity, Mobility, and Change


Sep 07, 2019
Lav R. Varshney, Nitish Shirish Keskar, Richard Socher

Figure 1 for Pretrained AI Models: Performativity, Mobility, and Change
Figure 2 for Pretrained AI Models: Performativity, Mobility, and Change
Figure 3 for Pretrained AI Models: Performativity, Mobility, and Change

  Access Paper or Ask Questions

Neural Text Summarization: A Critical Evaluation


Aug 23, 2019
Wojciech KryΕ›ciΕ„ski, Nitish Shirish Keskar, Bryan McCann, Caiming Xiong, Richard Socher

Figure 1 for Neural Text Summarization: A Critical Evaluation
Figure 2 for Neural Text Summarization: A Critical Evaluation
Figure 3 for Neural Text Summarization: A Critical Evaluation
Figure 4 for Neural Text Summarization: A Critical Evaluation
* To appear in EMNLP 2019, 13 pages, 2 figures, 6 tables 
  Access Paper or Ask Questions

XLDA: Cross-Lingual Data Augmentation for Natural Language Inference and Question Answering


May 27, 2019
Jasdeep Singh, Bryan McCann, Nitish Shirish Keskar, Caiming Xiong, Richard Socher

Figure 1 for XLDA: Cross-Lingual Data Augmentation for Natural Language Inference and Question Answering
Figure 2 for XLDA: Cross-Lingual Data Augmentation for Natural Language Inference and Question Answering
Figure 3 for XLDA: Cross-Lingual Data Augmentation for Natural Language Inference and Question Answering
Figure 4 for XLDA: Cross-Lingual Data Augmentation for Natural Language Inference and Question Answering

  Access Paper or Ask Questions

Unifying Question Answering and Text Classification via Span Extraction


Apr 19, 2019
Nitish Shirish Keskar, Bryan McCann, Caiming Xiong, Richard Socher

Figure 1 for Unifying Question Answering and Text Classification via Span Extraction
Figure 2 for Unifying Question Answering and Text Classification via Span Extraction
Figure 3 for Unifying Question Answering and Text Classification via Span Extraction
Figure 4 for Unifying Question Answering and Text Classification via Span Extraction

  Access Paper or Ask Questions

Coarse-grain Fine-grain Coattention Network for Multi-evidence Question Answering


Jan 03, 2019
Victor Zhong, Caiming Xiong, Nitish Shirish Keskar, Richard Socher

Figure 1 for Coarse-grain Fine-grain Coattention Network for Multi-evidence Question Answering
Figure 2 for Coarse-grain Fine-grain Coattention Network for Multi-evidence Question Answering
Figure 3 for Coarse-grain Fine-grain Coattention Network for Multi-evidence Question Answering
Figure 4 for Coarse-grain Fine-grain Coattention Network for Multi-evidence Question Answering
* ICLR 2019; 9 pages, 7 figures 
  Access Paper or Ask Questions

A Closer Look at Deep Learning Heuristics: Learning rate restarts, Warmup and Distillation


Oct 29, 2018
Akhilesh Gotmare, Nitish Shirish Keskar, Caiming Xiong, Richard Socher

Figure 1 for A Closer Look at Deep Learning Heuristics: Learning rate restarts, Warmup and Distillation
Figure 2 for A Closer Look at Deep Learning Heuristics: Learning rate restarts, Warmup and Distillation
Figure 3 for A Closer Look at Deep Learning Heuristics: Learning rate restarts, Warmup and Distillation
Figure 4 for A Closer Look at Deep Learning Heuristics: Learning rate restarts, Warmup and Distillation
* We use empirical tools of mode connectivity and SVCCA to investigate neural network training heuristics of learning rate restarts, warmup and knowledge distillation. arXiv admin note: text overlap with arXiv:1806.06977 
  Access Paper or Ask Questions

Identifying Generalization Properties in Neural Networks


Sep 19, 2018
Huan Wang, Nitish Shirish Keskar, Caiming Xiong, Richard Socher

Figure 1 for Identifying Generalization Properties in Neural Networks
Figure 2 for Identifying Generalization Properties in Neural Networks
Figure 3 for Identifying Generalization Properties in Neural Networks
Figure 4 for Identifying Generalization Properties in Neural Networks
* 23 pages 
  Access Paper or Ask Questions

The Natural Language Decathlon: Multitask Learning as Question Answering


Jun 20, 2018
Bryan McCann, Nitish Shirish Keskar, Caiming Xiong, Richard Socher

Figure 1 for The Natural Language Decathlon: Multitask Learning as Question Answering
Figure 2 for The Natural Language Decathlon: Multitask Learning as Question Answering
Figure 3 for The Natural Language Decathlon: Multitask Learning as Question Answering
Figure 4 for The Natural Language Decathlon: Multitask Learning as Question Answering

  Access Paper or Ask Questions

Using Mode Connectivity for Loss Landscape Analysis


Jun 18, 2018
Akhilesh Gotmare, Nitish Shirish Keskar, Caiming Xiong, Richard Socher

Figure 1 for Using Mode Connectivity for Loss Landscape Analysis
Figure 2 for Using Mode Connectivity for Loss Landscape Analysis
Figure 3 for Using Mode Connectivity for Loss Landscape Analysis
Figure 4 for Using Mode Connectivity for Loss Landscape Analysis
* Accepted as a workshop paper at ICML's Workshop on Modern Trends in Nonconvex Optimization for Machine Learning, 2018 
  Access Paper or Ask Questions

An Analysis of Neural Language Modeling at Multiple Scales


Mar 22, 2018
Stephen Merity, Nitish Shirish Keskar, Richard Socher

Figure 1 for An Analysis of Neural Language Modeling at Multiple Scales
Figure 2 for An Analysis of Neural Language Modeling at Multiple Scales
Figure 3 for An Analysis of Neural Language Modeling at Multiple Scales
Figure 4 for An Analysis of Neural Language Modeling at Multiple Scales

  Access Paper or Ask Questions

Improving Generalization Performance by Switching from Adam to SGD


Dec 20, 2017
Nitish Shirish Keskar, Richard Socher

Figure 1 for Improving Generalization Performance by Switching from Adam to SGD
Figure 2 for Improving Generalization Performance by Switching from Adam to SGD
Figure 3 for Improving Generalization Performance by Switching from Adam to SGD
Figure 4 for Improving Generalization Performance by Switching from Adam to SGD

  Access Paper or Ask Questions

Weighted Transformer Network for Machine Translation


Nov 06, 2017
Karim Ahmed, Nitish Shirish Keskar, Richard Socher

Figure 1 for Weighted Transformer Network for Machine Translation
Figure 2 for Weighted Transformer Network for Machine Translation
Figure 3 for Weighted Transformer Network for Machine Translation
Figure 4 for Weighted Transformer Network for Machine Translation

  Access Paper or Ask Questions

Regularizing and Optimizing LSTM Language Models


Aug 07, 2017
Stephen Merity, Nitish Shirish Keskar, Richard Socher

Figure 1 for Regularizing and Optimizing LSTM Language Models
Figure 2 for Regularizing and Optimizing LSTM Language Models
Figure 3 for Regularizing and Optimizing LSTM Language Models
Figure 4 for Regularizing and Optimizing LSTM Language Models

  Access Paper or Ask Questions

On Large-Batch Training for Deep Learning: Generalization Gap and Sharp Minima


Feb 09, 2017
Nitish Shirish Keskar, Dheevatsa Mudigere, Jorge Nocedal, Mikhail Smelyanskiy, Ping Tak Peter Tang

Figure 1 for On Large-Batch Training for Deep Learning: Generalization Gap and Sharp Minima
Figure 2 for On Large-Batch Training for Deep Learning: Generalization Gap and Sharp Minima
Figure 3 for On Large-Batch Training for Deep Learning: Generalization Gap and Sharp Minima
Figure 4 for On Large-Batch Training for Deep Learning: Generalization Gap and Sharp Minima
* Accepted as a conference paper at ICLR 2017 
  Access Paper or Ask Questions

adaQN: An Adaptive Quasi-Newton Algorithm for Training RNNs


Feb 23, 2016
Nitish Shirish Keskar, Albert S. Berahas

Figure 1 for adaQN: An Adaptive Quasi-Newton Algorithm for Training RNNs
Figure 2 for adaQN: An Adaptive Quasi-Newton Algorithm for Training RNNs

  Access Paper or Ask Questions