Posts by Collection

portfolio

publications

Decisions under drift: Adapting binary decision thresholds to drifts in test distribution

Published in In Proceedings of the 6th IBM Collaborative Academia Research Exchange Conference. ACM, New York, NY, USA, Article 17, 4 pages. DOI=http://dx.doi.org/10.1145/2662117.2662134, 2014

Recommended citation: Sachin Kumar, Vikas C. Raykar, and Priyanka Agrawal. (2014). Decisions under drift: Adapting binary decision thresholds to drifts in test distribution. In Proceedings of the 6th IBM Collaborative Academia Research Exchange Conference. ACM, New York, NY, USA, Article 17, 4 pages. DOI=http://dx.doi.org/10.1145/2662117.2662134.
Download Paper

Machine Translation with Continuous Outputs

Published in ICML 2018 workshop on Theoretical Foundations and Applications of Deep Generative Models, 2018

Recommended citation: Sachin Kumar, Yulia Tsvetkov. (2018). Machine Translation with Continuous Outputs. ICML 2018 workshop on Theoretical Foundations and Applications of Deep Generative Models.

Mining & Summarizing E-petitions for Enhanced Understanding of Public Opinion

Published in In Proceedings of the International Conference on Information and Knowledge Management (CIKM) 2018, 2018

Recommended citation: Shreshtha Mundra*, Sachin Kumar*, Manjira Sinha, Sandya Mannarswamy. (2018). Mining & Summarizing E-petitions for Enhanced Understanding of Public Opinion. In Proceedings of the International Conference on Information and Knowledge Management (CIKM) 2018.
Download Paper

Topics to Avoid: Demoting Latent Confounds in Text Classification

Published in 2019 Conference on Empirical Methods in Natural Language Processing (EMNLP) 2019, 2019

Recommended citation: Sachin Kumar, Shuly Wintner, Noah A. Smith, Yulia Tsvetkov. (2019). Topics to Avoid: Demoting Latent Confounds in Text Classification. 2019 Conference on Empirical Methods in Natural Language Processing (EMNLP) 2019.
Download Paper

End-to-End Differentiable GANs for Text Generation

Published in I Can’t Believe It’s Not Better! Workshop at NeurIPS 2020 (PMLR 137), 2020

Recommended citation: Sachin Kumar, Yulia Tsvetkov. (2020). End-to-End Differentiable GANs for Text Generation. I Can’t Believe It’s Not Better! Workshop at NeurIPS 2020 (PMLR 137).
Download Paper

Machine Translation into Low-Resource Language Varieties

Published in In the proceedings of 2021 Conference on Association of Computational Linguistics (ACL), 2021

Recommended citation: Sachin Kumar, Antonios Anastasopoulos, Shuly Wintner, Yulia Tsvetkov. (2021). Machine Translation into Low-Resource Language Varieties. In the proceedings of 2021 Conference on Association of Computational Linguistics (ACL).
Download Paper

Gradient-based Constrained Sampling from Language Models

Published in 2022 Conference on Empirical Methods in Natural Language Processing (EMNLP 2022), 2022

Recommended citation: Sachin Kumar, Biswajit Paria, Yulia Tsvetkov. (2022). Gradient-based Constrained Sampling from Language Models. 2022 Conference on Empirical Methods in Natural Language Processing (EMNLP 2022).
Download Paper

Reference-Free Sentence Summarization with Sharper Controllability through Symbolic Knowledge Distillation

Published in 2022 Conference on Empirical Methods in Natural Language Processing (EMNLP 2022), 2022

Recommended citation: Melanie Sclar, Peter West, Sachin Kumar, Yulia Tsvetkov and Yejin Choi. (2022). Reference-Free Sentence Summarization with Sharper Controllability through Symbolic Knowledge Distillation. 2022 Conference on Empirical Methods in Natural Language Processing (EMNLP 2022).
Download Paper

Do All Languages Cost the Same? Tokenization in the Era of Commercial Language Models

Published in 2023 Conference on Empirical Methods in Natural Language Processing (EMNLP 2023), 2023

Recommended citation: Orevaoghene Ahia, Sachin Kumar, Hila Gonen, Jungo Kasai, David R. Mortensen, Noah A. Smith, Yulia Tsvetkov. (2023). Do All Languages Cost the Same? Tokenization in the Era of Commercial Language Models. 2023 Conference on Empirical Methods in Natural Language Processing (EMNLP 2023).
Download Paper

Language Generation Models Can Cause Harm: So What Can We Do About It? An Actionable Survey

Published in 2023 Conference of the European Chapter of the Association for Computational Linguistics (EACL 2023), 2023

Recommended citation: Sachin Kumar*, Vidhisha Balachandran*, Lucille Njoo, Antonios Anastasopoulos, Yulia Tsvetkov. (2023). Language Generation Models Can Cause Harm: So What Can We Do About It? An Actionable Survey. 2023 Conference of the European Chapter of the Association for Computational Linguistics (EACL 2023).
Download Paper

Minding Language Models’ Theory of Mind: A Plug-and-Play Multi-Character Belief Tracker

Published in 2023 Conference of the Association for Computational Linguistics (ACL 2023). Outstanding Paper Award, 2023

Recommended citation: Melanie Sclar, Sachin Kumar, Peter West, Alane Suhr, Yejin Choi and Yulia Tsvetkov. (2023). Minding Language Models’ Theory of Mind: A Plug-and-Play Multi-Character Belief Tracker. 2023 Conference of the Association for Computational Linguistics (ACL 2023). Outstanding Paper Award.
Download Paper

On the Blind Spots of Model-Based Evaluation Metrics for Text Generation

Published in 2023 Conference of the Association for Computational Linguistics (ACL 2023), 2023

Recommended citation: Tianxing He, Jingyu Zhang, Tianle Wang, Sachin Kumar, Kyunghyun Cho, James Glass, Yulia Tsvetkov. (2023). On the Blind Spots of Model-Based Evaluation Metrics for Text Generation. 2023 Conference of the Association for Computational Linguistics (ACL 2023).
Download Paper

Dolma: an Open Corpus of Three Trillion Tokens for Language Model Pretraining Research

Published in 2024 Conference of the Association for Computational Linguistics (ACL 2024). Best Resource Paper Award, 2024

Recommended citation: Luca Soldaini, Rodney Kinney, Akshita Bhagia, Dustin Schwenk, David Atkinson, Russell Authur, Ben Bogin, Khyathi Chandu, Jennifer Dumas, Yanai Elazar, Valentin Hofmann, Ananya Harsh Jha, Sachin Kumar, Li Lucy, Xinxi Lyu, Nathan Lambert, Ian Magnusson, Jacob Morrison, Niklas Muennighoff, Aakanksha Naik, Crystal Nam, Matthew E. Peters, Abhilasha Ravichander, Kyle Richardson, Zejiang Shen, Emma Strubell, Nishant Subramani, Oyvind Tafjord, Pete Walsh, Luke Zettlemoyer, Noah A. Smith, Hannaneh Hajishirzi, Iz Beltagy, Dirk Groeneveld, Jesse Dodge, Kyle Lo. (2024). Dolma: an Open Corpus of Three Trillion Tokens for Language Model Pretraining Research. 2024 Conference of the Association for Computational Linguistics (ACL 2024). Best Resource Paper Award.
Download Paper

MAGNET: Improving the Multilingual Fairness of Language Models with Adaptive Gradient-Based Tokenization

Published in Thirty-eighth Conference on Neural Information Processing Systems (NeurIPS) 2024, 2024

Recommended citation: Orevaoghene Ahia, Sachin Kumar, Hila Gonen, Valentin Hofmann, Tomasz Limisiewicz, Yulia Tsvetkov, Noah A. Smith. (2024). MAGNET: Improving the Multilingual Fairness of Language Models with Adaptive Gradient-Based Tokenization. Thirty-eighth Conference on Neural Information Processing Systems (NeurIPS) 2024.
Download Paper

SSD-2: Scaling and Inference-time Fusion of Diffusion Language Models

Published in 2024 Conference of the North American Chapter of the Association for Computational Linguistics (NAACL 2024), 2024

Recommended citation: Xiaochuang Han, Sachin Kumar, Yulia Tsvetkov, Marjan Ghazvininejad. (2024). SSD-2: Scaling and Inference-time Fusion of Diffusion Language Models. 2024 Conference of the North American Chapter of the Association for Computational Linguistics (NAACL 2024).
Download Paper

The Art of Saying No: Contextual Noncompliance in Language Models

Published in Thirty-eighth Conference on Neural Information Processing Systems (NeurIPS) 2024: Datasets and Benchmarks, 2024

Recommended citation: Faeze Brahman*, Sachin Kumar*, Vidhisha Balachandran, Pradeep Dasigi, Valentina Pyatkin, Abhilasha Ravichander, Sarah Wiegreffe, Nouha Dziri, Khyathi Chandu, Jack Hessel, Yulia Tsvetkov, Noah A. Smith, Yejin Choi, Hannaneh Hajishirzi. (2024). The Art of Saying No: Contextual Noncompliance in Language Models. Thirty-eighth Conference on Neural Information Processing Systems (NeurIPS) 2024: Datasets and Benchmarks.
Download Paper

What Constitutes a Faithful Summary? Preserving Author Perspectives in News Summarization

Published in 2024 Conference of the North American Chapter of the Association for Computational Linguistics (NAACL 2024), 2024

Recommended citation: YuHan Liu, Shangbin Feng, Xiaochuang Han, Vidhisha Balachandran, Chan Young Park, Sachin Kumar, Yulia Tsvetkov. (2024). What Constitutes a Faithful Summary? Preserving Author Perspectives in News Summarization. 2024 Conference of the North American Chapter of the Association for Computational Linguistics (NAACL 2024).
Download Paper

WildTeaming at Scale: From In-the-Wild Jailbreaks to (Adversarially) Safer Language Models

Published in Thirty-eighth Conference on Neural Information Processing Systems (NeurIPS) 2024, 2024

Recommended citation: Liwei Jiang, Kavel Rao, Seungju Han, Allyson Ettinger, Faeze Brahman, Sachin Kumar, Niloofar Mireshghallah, Ximing Lu, Maarten Sap, Yejin Choi, Nouha Dziri. (2024). WildTeaming at Scale: From In-the-Wild Jailbreaks to (Adversarially) Safer Language Models. Thirty-eighth Conference on Neural Information Processing Systems (NeurIPS) 2024.
Download Paper

A Survey on Personalized and Pluralistic Preference Alignment in Large Language Models

Published in Conference on Language Modeling (COLM) 2025, 2025

Recommended citation: Zhouhang Xie, Junda Wu, Yiran Shen, Yu Xia, Xintong Li, Aaron Chang, Ryan Rossi, Sachin Kumar, Bodhisattwa Prasad Majumder, Jingbo Shang, Prithviraj Ammanabrolu, Julian McAuley. (2025). A Survey on Personalized and Pluralistic Preference Alignment in Large Language Models. Conference on Language Modeling (COLM) 2025.
Download Paper

BLAB: Brutally Long Audio Bench

Published in preprint, 2025

Recommended citation: Orevaoghene Ahia, Martijn Bartelds, Kabir Ahuja, Hila Gonen, Valentin Hofmann, Siddhant Arora, Shuyue Stella Li, Vishal Puttagunta, Mofetoluwa Adeyemi, Charishma Buchireddy, Ben Walls, Noah Bennett, Shinji Watanabe, Noah A. Smith, Yulia Tsvetkov, Sachin Kumar. (2025). BLAB: Brutally Long Audio Bench. preprint.
Download Paper

ComPO: Community Preferences for Language Model Personalization

Published in 2025 Annual Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics (NAACL 2025). *=equal contribution, 2025

Recommended citation: Sachin Kumar*, Chan Young Park*, Yulia Tsvetkov, Noah A. Smith, Hannaneh Hajishirzi. (2025). ComPO: Community Preferences for Language Model Personalization. 2025 Annual Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics (NAACL 2025). *=equal contribution.
Download Paper

GroundCocoa: A Benchmark for Evaluating Compositional & Conditional Reasoning in Language Models

Published in 2025 Annual Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics (NAACL 2025), 2025

Recommended citation: Harsh Kohli, Sachin Kumar, Huan Sun. (2025). GroundCocoa: A Benchmark for Evaluating Compositional & Conditional Reasoning in Language Models. 2025 Annual Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics (NAACL 2025).
Download Paper

Hybrid Preferences: Learning to Route Instances for Human vs. AI Feedback

Published in 2025 Conference of the Association for Computational Linguistics (ACL 2025), 2025

Recommended citation: Lester James V. Miranda, Yizhong Wang, Yanai Elazar, Sachin Kumar, Valentina Pyatkin, Faeze Brahman, Noah A. Smith, Hannaneh Hajishirzi, Pradeep Dasigi. (2025). Hybrid Preferences: Learning to Route Instances for Human vs. AI Feedback. 2025 Conference of the Association for Computational Linguistics (ACL 2025).
Download Paper

RewardBench: Evaluating Reward Models for Language Modeling

Published in 2025 Annual Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics (NAACL 2025) Findings, 2025

Recommended citation: Nathan Lambert, Valentina Pyatkin, Jacob Morrison, LJ Miranda, Bill Yuchen Lin, Khyathi Chandu, Nouha Dziri, Sachin Kumar, Tom Zick, Yejin Choi, Noah A. Smith, Hannaneh Hajishirzi. (2025). RewardBench: Evaluating Reward Models for Language Modeling. 2025 Annual Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics (NAACL 2025) Findings.
Download Paper

ScholarEval: Research Idea Evaluation Grounded in Literature

Published in preprint. *=equal contribution, 2025

Recommended citation: Hanane Nour Moussa*, Patrick Queiroz Da Silva*, Daniel Adu-Ampratwum, Alyson East, Zitong Lu, Nikki Puccetti, Mingyi Xue, Huan Sun, Bodhisattwa Prasad Majumder, Sachin Kumar. (2025). ScholarEval: Research Idea Evaluation Grounded in Literature. preprint. *=equal contribution.
Download Paper

Steering off Course: Reliability Challenges in Steering Language Models

Published in 2025 Conference of the Association for Computational Linguistics (ACL 2025). Oral (top 8%), Panel (top 0.8%), 2025

Recommended citation: Patrick Queiroz Da Silva, Hari Sethuraman, Dheeraj Rajagopal, Hannaneh Hajishirzi, Sachin Kumar. (2025). Steering off Course: Reliability Challenges in Steering Language Models. 2025 Conference of the Association for Computational Linguistics (ACL 2025). Oral (top 8%), Panel (top 0.8%).
Download Paper

Continually Adding New Languages to Multilingual Language Models

Published in Transactions on Machine Learning Research (TMLR), 2026; also presented at Multilingual and Equitable Language Technologies (MELT) Workshop @COLM 2025, 2026

Recommended citation: Abraham Toluwase Owodunni, Sachin Kumar. (2026). Continually Adding New Languages to Multilingual Language Models. Transactions on Machine Learning Research (TMLR), 2026; also presented at Multilingual and Equitable Language Technologies (MELT) Workshop @COLM 2025.
Download Paper

FLEXITOKENS: Flexible Tokenization for Evolving Language Models

Published in Findings of the Association for Computational Linguistics: ACL 2026; also presented at Tokenization Workshop @ICML 2025, 2026

Recommended citation: Abraham Toluwase Owodunni, Orevaoghene Ahia, Sachin Kumar. (2026). FLEXITOKENS: Flexible Tokenization for Evolving Language Models. Findings of the Association for Computational Linguistics: ACL 2026; also presented at Tokenization Workshop @ICML 2025.
Download Paper

Reasoning Up the Instruction Ladder for Controllable Language Models

Published in Findings of the Association for Computational Linguistics: ACL 2026; also presented at Foundations of Reasoning in Language Models (FoRLM) Workshop at NeurIPS 2025, 2026

Recommended citation: Zishuo Zheng, Vidhisha Balachandran, Chan Young Park, Faeze Brahman, Sachin Kumar. (2026). Reasoning Up the Instruction Ladder for Controllable Language Models. Findings of the Association for Computational Linguistics: ACL 2026; also presented at Foundations of Reasoning in Language Models (FoRLM) Workshop at NeurIPS 2025.
Download Paper

Position: We Need A Unified Definition of Hallucination (It’s The World Model, Stupid!)

Published in International Conference on Machine Learning (ICML 2026), Position Paper Track, 2026

Recommended citation: Emmy Liu, Varun Prashant Gangal, Chelsea Zou, Michael Yu, Xiaoqi Huang, Alex Chang, Zhuofu Tao, Karanpartap Singh, Sachin Kumar, Steven Y. Feng. (2026). Position: We Need A Unified Definition of Hallucination (It’s The World Model, Stupid!). International Conference on Machine Learning (ICML 2026), Position Paper Track.
Download Paper

Accepted with Minor Revisions: Value of AI-Assisted Scientific Writing

Published in 2026 ACM Conference on Intelligent User Interfaces (IUI 2026). *=equal contribution, 2026

Recommended citation: Sanchaita Hazra*, Doeun Lee*, Bodhisattwa Prasad Majumder, Sachin Kumar. (2026). Accepted with Minor Revisions: Value of AI-Assisted Scientific Writing. 2026 ACM Conference on Intelligent User Interfaces (IUI 2026). *=equal contribution.
Download Paper

Connecting the Dots: Enduring Performance Gaps in ASR Systems Between Standard and Minority English Varieties

Published in 2026 IEEE Spoken Language Technology (SLT 2026). To Appear, 2026

Recommended citation: Alexander Metzger, Aruna Srivastava, Ruslan Mukhamedvaleev, Eunjung Yeo, Syed Ishtiaque Ahmed, Nina Markl, Sachin Kumar, Farhan Samir. (2026). Connecting the Dots: Enduring Performance Gaps in ASR Systems Between Standard and Minority English Varieties. 2026 IEEE Spoken Language Technology (SLT 2026). To Appear.
Download Paper

One Form to Transfer Them All: Pretraining Multilingual Language Models Beyond Native Orthography

Published in Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026), 2026

Recommended citation: Muge Zhang, Aaron Jencks, Krishna Badikela, Yulia Tsvetkov, Sachin Kumar. (2026). One Form to Transfer Them All: Pretraining Multilingual Language Models Beyond Native Orthography. Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026).
Download Paper

Process-Oriented Evaluation of AI-Assisted Scientific Writing

Published in Third Conference on Language Modeling (COLM 2026), 2026

Recommended citation: Patrick Queiroz Da Silva, Sanchaita Hazra, Doeun Lee, Sachin Kumar, Bodhisattwa Prasad Majumder. (2026). Process-Oriented Evaluation of AI-Assisted Scientific Writing. Third Conference on Language Modeling (COLM 2026).
Download Paper

The Cost of Sounding Different: Accent Bias in Audio Language Models

Published in Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026). Outstanding Paper Award 🏆, 2026

Recommended citation: Aruna Srivastava, Orevaoghene Ahia, Li Lucy, Ashley Christendat, Sameep Chattopadhyay, Farhan Samir, Tejumade Afonja, Valentin Hofmann, Sachin Kumar, Noah A. Smith, Yulia Tsvetkov. (2026). The Cost of Sounding Different: Accent Bias in Audio Language Models. Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026). Outstanding Paper Award 🏆.
Download Paper

Tokenization: A Survey for Modern NLP

Published in preprint, 2026

Recommended citation: Marco Cognetta, Christopher Akiki, Pawan Sasanka Ammanamanchi, Catherine Arnett, Thomas Bauwens, Pavel Chizhov, Pieter Delobelle, Konstantin Dobler, Desmond Elliott, Omer Goldman, Tatsuya Hiraoka, Valentin Hofmann, Eugene Jang, Julie Kallini, Ilker Kesen, Sachin Kumar, Sander Land, Jindřich Libovický, Tomasz Limisiewicz, Alisa Liu, Clara Meister, Benjamin Minixhofer, Sangwhan Moon, Tiago Pimentel, David Pohl, Suchir Salhan, Tomohiro Sawada, Pavel Stepachev, Gianluca Vico, Philip Whittington, Xiulin Yang, Yuval Pinter. (2026). Tokenization: A Survey for Modern NLP. preprint.
Download Paper

When Cases Get Rare: A Retrieval Benchmark for Off-Guideline Clinical Question Answering

Published in ICML 2026 Workshop on Structured Data for Health (SD4H). Poster, 2026

Recommended citation: Doeun Lee, Muge Zhang, Yi Yu, Ashish Manne, Stephen Koesters, Frank Wen, Brady Buchanan, Lynda Villagomez, Oluwatoba Moninuola, James Lim, Kathryn Tobin, Andrew Srisuwananukorn, Ping Zhang, Sachin Kumar. (2026). When Cases Get Rare: A Retrieval Benchmark for Off-Guideline Clinical Question Answering. ICML 2026 Workshop on Structured Data for Health (SD4H). Poster.
Download Paper

talks

teaching