Publications

2026

1. NeurIPS

   Efficient Forecasting of Task Failures in LLM Agents through Adaptive Fault Injection

   Vartika Sengar, Parth Thakkar, Pranoy Panda, Shrey Satapara, Emmy Liu, and 4 more authors

   In Advances in Neural Information Processing Systems (NeurIPS 2026), 2026

   Earlier version presented at Forecast@ICML 2026.

   BibTeX Workshop version Workshop PDF

   @inproceedings{task-failure-forecasting,
     title = {Efficient Forecasting of Task Failures in LLM Agents through Adaptive Fault Injection},
     author = {Sengar, Vartika and Thakkar, Parth and Panda, Pranoy and Satapara, Shrey and Liu, Emmy and Viswanathan, Vijay and Takemori, Sho and Neubig, Graham and Devaguptapu, Chaitanya},
     booktitle = {Advances in Neural Information Processing Systems (NeurIPS 2026)},
     year = {2026},
     note = {Earlier version presented at Forecast@ICML 2026.},
     keywords = {conference},
   }

2. NeurIPS

   HalluWorld: A Controlled Benchmark for Hallucination via Reference World Models

   Emmy Liu, Varun Prashant Gangal, Michael Yu, Zhuofu Tao, Karanpartap Singh, and 2 more authors

   In NeurIPS 2026 Evaluations and Datasets Track, 2026

   arXiv BibTeX PDF

   @inproceedings{halluworld,
     title = {HalluWorld: A Controlled Benchmark for Hallucination via Reference World Models},
     author = {Liu, Emmy and Gangal, Varun Prashant and Yu, Michael and Tao, Zhuofu and Singh, Karanpartap and Kumar, Sachin and Feng, Steven Y.},
     booktitle = {NeurIPS 2026 Evaluations and Datasets Track},
     year = {2026},
     archiveprefix = {arXiv},
     primaryclass = {cs.CL},
     keywords = {conference},
   }

3. COLM

   What Do Language Models Learn and When? The Implicit Curriculum Hypothesis

   Emmy Liu, Kaiser Sun, Millicent Li, Isabelle Lee, Lindia Tjuatja, and 2 more authors

   In Proceedings of the 3rd Conference on Language Modeling (COLM 2026), 2026

   arXiv BibTeX PDF Code

   @inproceedings{implicit-curriculum,
     title = {What Do Language Models Learn and When? The Implicit Curriculum Hypothesis},
     author = {Liu, Emmy and Sun, Kaiser and Li, Millicent and Lee, Isabelle and Tjuatja, Lindia and Huang, Jen-tse and Neubig, Graham},
     booktitle = {Proceedings of the 3rd Conference on Language Modeling (COLM 2026)},
     year = {2026},
     keywords = {conference},
   }

4. ICML

   Midtraining Bridges Pretraining and Posttraining Distributions

   Emmy Liu, Chenyan Xiong, and Graham Neubig

   In Proceedings of the 43rd International Conference on Machine Learning (ICML 2026), 2026

   arXiv BibTeX PDF Code

   @inproceedings{midtraining,
     title = {Midtraining Bridges Pretraining and Posttraining Distributions},
     author = {Liu, Emmy and Xiong, Chenyan and Neubig, Graham},
     year = {2026},
     booktitle = {Proceedings of the 43rd International Conference on Machine Learning (ICML 2026)},
     keyword = {conference},
   }

5. ICML

   Hybrid-Gym: Training Coding Agents to Generalize Across Tasks

   Yiqing Xie, Emmy Liu, Gaokai Zhang, Nachiket Kotalwar, Shubham Gandhi, and 5 more authors

   In Proceedings of the 43rd International Conference on Machine Learning (ICML 2026), 2026

   arXiv BibTeX PDF Code

   @inproceedings{hybridgym,
     title = {Hybrid-Gym: Training Coding Agents to Generalize Across Tasks},
     author = {Xie, Yiqing and Liu, Emmy and Zhang, Gaokai and Kotalwar, Nachiket and Gandhi, Shubham and Acharya, Sathwik and Wang, Xingyao and Rose, Carolyn and Neubig, Graham and Fried, Daniel},
     year = {2026},
     booktitle = {Proceedings of the 43rd International Conference on Machine Learning (ICML 2026)},
     keyword = {conference},
   }

6. ICML Position

   A Unified Definition of Hallucination: It’s The World Model, Stupid!

   Emmy Liu, Varun Gangal, Chelsea Zou, Michael Yu, Xiaoqi Huang, and 5 more authors

   In Proceedings of the 43rd International Conference on Machine Learning, Position Paper Track (ICML 2026), 2026

   arXiv BibTeX PDF Code

   @inproceedings{hallucination-worldmodel,
     title = {A Unified Definition of Hallucination: It's The World Model, Stupid!},
     author = {Liu, Emmy and Gangal, Varun and Zou, Chelsea and Yu, Michael and Huang, Xiaoqi and Chang, Alex and Tao, Zhuofu and Singh, Karan and Kumar, Sachin and Feng, Steven Y.},
     booktitle = {Proceedings of the 43rd International Conference on Machine Learning, Position Paper Track (ICML 2026)},
     year = {2026},
     keywords = {conference, position paper},
   }

1. EvalEval@ACL

   Rigorous Interpretation Is a Form of Evaluation

   Isabelle Lee, Emmy Liu, Cathy Jiao, Brihi Joshi, Dani Yogatama, and 2 more authors

   In Proceedings of the ACL 2026 Workshop on Evaluating Evaluations (EvalEval), 2026

   arXiv HTML PDF

1. arXiv

   To Memorize or to Retrieve: Scaling the Interaction Between Pretraining and Retrieval

   Karan Singh, Michael Yu, Varun Gangal, Zhuofu Tao, Sachin Kumar, and 2 more authors

   2026

   arXiv BibTeX PDF Code

   @misc{memorize-or-retrieve,
     title = {To Memorize or to Retrieve: Scaling the Interaction Between Pretraining and Retrieval},
     author = {Singh, Karan and Yu, Michael and Gangal, Varun and Tao, Zhuofu and Kumar, Sachin and Liu, Emmy and Feng, Steven Y.},
     year = {2026},
     archiveprefix = {arXiv},
     primaryclass = {cs.CL},
     keywords = {preprint},
   }

2. arXiv

   Egocentric Bias in Vision-Language Models

   Maijunxian Wang, Yijiang Li, Bingyang Wang, Tianwei Zhao, Ran Ji, and 4 more authors

   2026

   arXiv PDF

2025

1. EMNLP

   Not-Just-Scaling Laws: Towards a Better Understanding of the Downstream Impact of Language Model Design Decisions

   Emmy Liu, Amanda Bertsch, Lintang Sutawika, Lindia Tjuatja, Patrick Fernandes, and 7 more authors

   In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing (EMNLP 2025), 2025

   arXiv BibTeX PDF

   @inproceedings{pbj,
     title = {Not-Just-Scaling Laws: Towards a Better Understanding of the Downstream Impact of Language Model Design Decisions},
     author = {Liu, Emmy and Bertsch, Amanda and Sutawika, Lintang and Tjuatja, Lindia and Fernandes, Patrick and Marinov, Lara and Chen, Michael and Singhal, Shreya and Lawrence, Carolin and Raghunathan, Aditi and Gashteovski, Kiril and Neubig, Graham},
     booktitle = {Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing (EMNLP 2025)},
     year = {2025},
     publisher = {Association for Computational Linguistics},
     keywords = {conference},
   }

2. NAACL Findings

   Language Modeling with Editable External Knowledge

   Belinda Z. Li, Emmy Liu, Alexis Ross, Abbas Zeitoun, Graham Neubig, and 1 more author

   In Findings of the Association for Computational Linguistics: NAACL 2025, 2025

   arXiv BibTeX PDF Code

   @inproceedings{mem-lm,
     title = {Language Modeling with Editable External Knowledge},
     author = {Li, Belinda Z. and Liu, Emmy and Ross, Alexis and Zeitoun, Abbas and Neubig, Graham and Andreas, Jacob},
     year = {2025},
     booktitle = {Findings of the Association for Computational Linguistics: NAACL 2025},
   }

1. arXiv

   Datasheets Aren’t Enough: DataRubrics for Automated Quality Metrics and Accountability

   Genta Indra Winata*, David Anugraha*, Emmy Liu*, Alham Fikri Aji*, Shou-Yi Hung, and 15 more authors

   2025

   arXiv Code

1. Thesis Proposal

   On the Developmental Science of Language Models

   Emmy Liu

   2025

   PhD Thesis Proposal, Carnegie Mellon University

   BibTeX PDF

   @misc{thesis-proposal,
     title = {On the Developmental Science of Language Models},
     author = {Liu, Emmy},
     year = {2025},
     note = {PhD Thesis Proposal, Carnegie Mellon University},
     keywords = {thesis}
   }

2024

1. COLM

   An Incomplete Loop: Deductive, Inductive, and Abductive Reasoning in Language Models

   Emmy Liu, Graham Neubig, and Jacob Andreas

   In Proceedings of the 1st Conference on Language Modeling, 2024

   arXiv BibTeX HTML PDF Code

   @inproceedings{abductive-reasoning,
     title = {An Incomplete Loop: Deductive, Inductive, and Abductive Reasoning in Language Models},
     author = {Liu, Emmy and Neubig, Graham and Andreas, Jacob},
     year = {2024},
     booktitle = {Proceedings of the 1st Conference on Language Modeling},
   }

2. Neurips

   Divergences between Language Models and Human Brains

   Yuchen Zhou, Emmy Liu, Graham Neubig, Michael J. Tarr, and Leila Wehbe

   In Advances in Neural Information Processing Systems 37, 2024

   arXiv BibTeX PDF

   @inproceedings{brain-lm,
     title = {Divergences between Language Models and Human Brains},
     author = {Zhou, Yuchen and Liu, Emmy and Neubig, Graham and Tarr, Michael J. and Wehbe, Leila},
     year = {2024},
     booktitle = {Advances in Neural Information Processing Systems 37},
     address = {Vancouver, Canada},
   }

3. NAACL

   Program-Aided Reasoners (Better) Know What They Know

   Anubha Kabra, Sanketh Rangreji, Yash Mathur, Aman Madaan, Emmy Liu, and 1 more author

   In Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, 2024

   arXiv BibTeX PDF

     title = {Program-Aided Reasoners (Better) Know What They Know},
     author = {Kabra, Anubha and Rangreji, Sanketh and Mathur, Yash and Madaan, Aman and Liu, Emmy and Neubig, Graham},
     year = {2024},
     booktitle = {Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies},
     address = {Mexico City, Mexico}
   }

2023

1. EMNLP

   Crossing the Threshold: Idiomatic Machine Translation through Retrieval Augmentation and Loss Weighting

   Emmy Liu, Aditi Chaudhary, and Graham Neubig

   In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing (EMNLP), 2023

   arXiv BibTeX PDF Code

   @inproceedings{idiom-translation,
     title = {Crossing the Threshold: Idiomatic Machine Translation through Retrieval Augmentation and Loss Weighting},
     author = {Liu, Emmy and Chaudhary, Aditi and Neubig, Graham},
     year = {2023},
     booktitle = {Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing (EMNLP)},
     address = {Singapore},
     keywords = {conference},
   }

2. TACL

   Bridging the Gap: A Survey on Integrating (Human) Feedback for Natural Language Generation

   Patrick Fernandes, Aman Madaan, Emmy Liu, António Farinhas, Pedro Henrique Martins, and 6 more authors

   Transactions of the Association for Computational Linguistics, 2023

   arXiv BibTeX PDF

   @article{hf_survey,
     title = {Bridging the Gap: A Survey on Integrating (Human) Feedback for Natural Language Generation},
     author = {Fernandes, Patrick and Madaan, Aman and Liu, Emmy and Farinhas, António and Martins, Pedro Henrique and Bertsch, Amanda and de Souza, José G. C. and Zhou, Shuyan and Wu, Tongshuang and Neubig, Graham and Martins, André F. T.},
     year = {2023},
     journal = {Transactions of the Association for Computational Linguistics},
     keywords = {journal},
   }

3. star_sem

   Syntax and Semantics Meet in the "Middle": Probing the Syntax-Semantics Interface of LMs Through Agentivity

   Lindia Tjuatja, Emmy Liu, Lori Levin, and Graham Neubig

   In Proceedings of the 12th Joint Conference on Lexical and Computational Semantics, 2023

   arXiv BibTeX PDF

   @inproceedings{lm-agentivity,
     title = {Syntax and Semantics Meet in the "Middle": Probing the Syntax-Semantics Interface of LMs Through Agentivity},
     author = {Tjuatja, Lindia and Liu, Emmy and Levin, Lori and Neubig, Graham},
     year = {2023},
     booktitle = {Proceedings of the 12th Joint Conference on Lexical and Computational Semantics},
   }

4. ACL

   When Does Translation Require Context? A Data-driven, Multilingual Exploration (Best Resource Paper)

   Patrick Fernandes, Kayo Yin, Emmy Liu, André F. T. Martins, and Graham Neubig

   In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics, 2023

   arXiv BibTeX PDF

   @inproceedings{muda,
     title = {When Does Translation Require Context? A Data-driven, Multilingual Exploration (Best Resource Paper)},
     author = {Fernandes, Patrick and Yin, Kayo and Liu, Emmy and Martins, André F. T. and Neubig, Graham},
     booktitle = {Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics},
     year = {2023},
   }

5. ACL Findings

   Multi-lingual and Multi-cultural Figurative Language Understanding

   Anubha Kabra*, Emmy Liu*, Simran Khanuja*, Alham Fikri Aji, Genta Indra Winata, and 4 more authors

   In Findings of the Association for Computational Linguistics: ACL 2023, 2023

   arXiv BibTeX PDF

   @inproceedings{multi-figqa,
     title = {Multi-lingual and Multi-cultural Figurative Language Understanding},
     author = {Kabra*, Anubha and Liu*, Emmy and Khanuja*, Simran and Aji, Alham Fikri and Winata, Genta Indra and Cahyawijaya, Samuel and Aremu, Anuoluwapo and Ogayo, Perez and Neubig, Graham},
     booktitle = {Findings of the Association for Computational Linguistics: ACL 2023},
     year = {2023},
   }

6. ICLR

   Computational Language Acquisition with Theory of Mind

   Andy Liu, Hao Zhu, Emmy Liu, Yonatan Bisk, and Graham Neubig

   In The Eleventh International Conference on Learning Representations, 2023

   arXiv BibTeX PDF

   @inproceedings{tom-learning,
     title = {Computational Language Acquisition with Theory of Mind},
     author = {Liu, Andy and Zhu, Hao and Liu, Emmy and Bisk, Yonatan and Neubig, Graham},
     booktitle = {The Eleventh International Conference on Learning Representations},
     year = {2023},
   }

7. CHI

   Exam Eustress: Designing a Brief Online Intervention for Helping Students Identify Positive Aspects of Stress

   Mohi Reza, Angela B. Zavaleta, Emmy Liu, Tong Li, Calista Barber, and 2 more authors

   In CHI ’23: CHI Conference on Human Factors in Computing Systems, 2023

   BibTeX PDF

   @inproceedings{exam-stress,
     title = {Exam Eustress: Designing a Brief Online Intervention for Helping Students Identify Positive Aspects of Stress},
     author = {Reza, Mohi and Zavaleta, Angela B. and Liu, Emmy and Li, Tong and Barber, Calista and Liang, Zhongyuan and Williams, Joseph Jay},
     year = {2023},
     booktitle = {CHI '23: CHI Conference on Human Factors in Computing Systems},
     doi = {10.1145/3544548.3581368},
     url = {https://doi.org/10.1145/3544548.3581368},
   }

2022

1. NAACL

   Testing the Ability of Language Models to Interpret Figurative Language

   Emmy Liu, Chenxuan Cui, Kenneth Zheng, and Graham Neubig

   In Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Jul 2022

   arXiv BibTeX PDF

   @inproceedings{liu-etal-2022-testing,
     title = {Testing the Ability of Language Models to Interpret Figurative Language},
     author = {Liu, Emmy and Cui, Chenxuan and Zheng, Kenneth and Neubig, Graham},
     booktitle = {Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies},
     month = jul,
     year = {2022},
     address = {Seattle, United States},
     publisher = {Association for Computational Linguistics},
     url = {https://aclanthology.org/2022.naacl-main.330},
     doi = {10.18653/v1/2022.naacl-main.330},
     pages = {4437--4452},
   }

2. EMNLP

   Are Representations Built from the Ground Up? An Empirical Examination of Local Composition in Language Models

   Emmy Liu, and Graham Neubig

   In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, Dec 2022

   arXiv BibTeX PDF

   @inproceedings{compositionality-22,
     author = {Liu, Emmy and Neubig, Graham},
     address = {"Abu Dhabi, United Arab Emirates"},
     month = dec,
     title = {Are Representations Built from the Ground Up? An Empirical Examination of Local Composition in Language Models},
     booktitle = {Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing},
     publisher = {Association for Computational Linguistics},
     year = {2022},
   }

3. CogSci

   The emergence of moral foundations in child language development

   Aida Ramezani, Emmy Liu, Renato Ferreira Pinto Junior, Spike W.S. Lee, and Yang Xu

   In Proceedings of the Annual Meeting of the Cognitive Science Society, Jul 2022

   BibTeX PDF

   @inproceedings{child-morality-22,
     author = {Ramezani, Aida and Liu, Emmy and Ferreira Pinto Junior, Renato and Lee, Spike W.S. and Xu, Yang},
     title = {The emergence of moral foundations in child language development},
     booktitle = {Proceedings of the Annual Meeting of the Cognitive Science Society},
     year = {2022},
     volume = {44},
     url = {https://escholarship.org/uc/item/3qv9h3j9},
     month = jul,
   }

4. PLoS One

   Chaining and the temporal dynamics of scientists’ publishing behaviour

   Emmy Liu, and Yang Xu

   PLoS One, Jul 2022

   BibTeX

   @article{plos-metascience,
     author = {Liu, Emmy and Xu, Yang},
     year = {2022},
     title = {Chaining and the temporal dynamics of scientists’ publishing behaviour},
     journal = {PLoS One},
     doi = {10.1371/journal.pone.0278389},
     url = {https://doi.org/10.1371/journal.pone.0278389},
   }

1. GeBNLP@NAACL

   Assessing Group-level Gender Bias in Professional Evaluations: The Case of Medical Student End-of-Shift Feedback

   Emmy Liu, Michael Henry Tessler, Nicole Dubosh, Katherine Hiller, and Roger Levy

   In Proceedings of the 4th Workshop on Gender Bias in Natural Language Processing (GeBNLP), Jul 2022

   arXiv BibTeX PDF

   @inproceedings{liu-etal-2022-assessing,
     title = {Assessing Group-level Gender Bias in Professional Evaluations: The Case of Medical Student End-of-Shift Feedback},
     author = {Liu, Emmy and Tessler, Michael Henry and Dubosh, Nicole and Hiller, Katherine and Levy, Roger},
     booktitle = {Proceedings of the 4th Workshop on Gender Bias in Natural Language Processing (GeBNLP)},
     month = jul,
     year = {2022},
     address = {Seattle, Washington},
     publisher = {Association for Computational Linguistics},
     url = {https://aclanthology.org/2022.gebnlp-1.11},
     doi = {10.18653/v1/2022.gebnlp-1.11},
     pages = {86--93},
   }

2. FigLang@EMNLP

   EUREKA: EUphemism Recognition Enhanced through Knn-based methods and Augmentation

   Sedrick Scott Keh, Rohit K. Bharadwaj, Emmy Liu, Simone Tedeschi, Varun Gangal, and 1 more author

   In Proceedings of the 3rd Workshop on Figurative Language Processing (FigLang), Dec 2022

   arXiv BibTeX PDF

   @inproceedings{euphemism-detect,
     author = {Keh, Sedrick Scott and Bharadwaj, Rohit K. and Liu, Emmy and Tedeschi, Simone and Gangal, Varun and Navigli, Roberto},
     title = {EUREKA: EUphemism Recognition Enhanced through Knn-based methods and Augmentation},
     booktitle = {Proceedings of the 3rd Workshop on Figurative Language Processing (FigLang)},
     publisher = {Association for Computational Linguistics},
     year = {2022},
     month = dec,
     copyright = {Creative Commons Attribution 4.0 International},
   }

2021

1. RL4ED@EDM

   Statistical Consequences of Duelling Bandits

   Nayan Saxena, Pan Chen, and Emmy Liu

   In EDM21 Workshop on Reinforcement Learning for Education, Dec 2021

   BibTeX

   @inproceedings{duelling-bandits,
     author = {Saxena, Nayan and Chen, Pan and Liu, Emmy},
     title = {Statistical Consequences of Duelling Bandits},
     year = {2021},
     booktitle = {EDM21 Workshop on Reinforcement Learning for Education},
   }

2020

1. CogSci

   Chaining and the Process of Scientific Innovation

   Emmy Liu, and Yang Xu

   In Proceedings of the Annual Meeting of the Cognitive Science Society, Jul 2020

   BibTeX

   @inproceedings{chaining-science,
     author = {Liu, Emmy and Xu, Yang},
     title = {Chaining and the Process of Scientific Innovation},
     year = {2020},
     booktitle = {Proceedings of the Annual Meeting of the Cognitive Science Society},
     volume = {42},
     pages = {1639--1646},
   }

2. Open Mind

   Numeral Systems across Languages Support Efficient Communication: From Approximate Numerosity to Recursion

   Emmy Liu, and Yang Xu

   Open Mind, Jul 2020

   BibTeX

   @article{numerals-journal,
     author = {Liu, Emmy and Xu, Yang},
     title = {Numeral Systems across Languages Support Efficient Communication: From Approximate Numerosity to Recursion},
     year = {2020},
     journal = {Open Mind},
     volume = {4},
     pages = {57--70},
   }

2019

1. CogSci

   Rapid information gain explains cross-linguistic tendenciees in numeral ordering

   Emmy Liu, and Yang Xu

   In Proceedings of the Annual Meeting of the Cognitive Science Society, Jul 2019

   BibTeX

   @inproceedings{numerals-cogsci,
     author = {Liu, Emmy and Xu, Yang},
     title = {Rapid information gain explains cross-linguistic tendenciees in numeral ordering},
     year = {2019},
     booktitle = {Proceedings of the Annual Meeting of the Cognitive Science Society},
     volume = {41},
     pages = {2166--2173},
   }

2. CUNY

   Ordering in numerals supports rapid information processing

   Emmy Liu, and Yang Xu

   In 32nd CUNY Conference on Human Sentence Processing, Jul 2019

   BibTeX

   @inproceedings{cuny,
     author = {Liu, Emmy and Xu, Yang},
     booktitle = {32nd CUNY Conference on Human Sentence Processing},
     year = {2019},
   }