publications

2024

1. arXiv

   HAICOSYSTEM: An Ecosystem for Sandboxing Safety Risks in Human-AI Interactions

   Xuhui Zhou, Hyunwoo Kim, Faeze Brahman, and 9 more authors

   2024

   Bib Paper Website

   @misc{zhou2024haicosystemecosystemsandboxingsafety,
     title = {HAICOSYSTEM: An Ecosystem for Sandboxing Safety Risks in Human-AI Interactions},
     author = {Zhou, Xuhui and Kim, Hyunwoo and Brahman, Faeze and Jiang, Liwei and Zhu, Hao and Lu, Ximing and Xu, Frank and Lin, Bill Yuchen and Choi, Yejin and Mireshghallah, Niloofar and Bras, Ronan Le and Sap, Maarten},
     year = {2024},
     eprint = {2409.16427},
     archiveprefix = {arXiv},
     booktitle = {arXiv},
     primaryclass = {cs.AI},
   }

2. arXiv

   AI-LieDar: Examine the Trade-off Between Utility and Truthfulness in LLM Agents

   Zhe Su, Xuhui Zhou, Sanketh Rangreji, and 4 more authors

   2024

   Bib Paper

   @misc{su2024ailiedarexaminetradeoffutility,
     title = {AI-LieDar: Examine the Trade-off Between Utility and Truthfulness in LLM Agents},
     author = {Su, Zhe and Zhou, Xuhui and Rangreji, Sanketh and Kabra, Anubha and Mendelsohn, Julia and Brahman, Faeze and Sap, Maarten},
     year = {2024},
     eprint = {2409.09013},
     archiveprefix = {arXiv},
     booktitle = {arXiv},
     primaryclass = {cs.AI},
   }

3. arXiv

   User-Driven Value Alignment: Understanding Users’ Perceptions and Strategies for Addressing Biased and Discriminatory Statements in AI Companions

   Xianzhe Fan, Qing Xiao, Xuhui Zhou, and 4 more authors

   2024

   Bib Paper

   @misc{fan2024userdrivenvaluealignmentunderstanding,
     title = {User-Driven Value Alignment: Understanding Users' Perceptions and Strategies for Addressing Biased and Discriminatory Statements in AI Companions},
     author = {Fan, Xianzhe and Xiao, Qing and Zhou, Xuhui and Pei, Jiaxin and Sap, Maarten and Lu, Zhicong and Shen, Hong},
     year = {2024},
     eprint = {2409.00862},
     archiveprefix = {arXiv},
     booktitle = {arXiv},
     primaryclass = {cs.HC},
   }

4. arXiv

   On the Resilience of Multi-Agent Systems with Malicious Agents

   Jen-tse Huang, Jiaxu Zhou, Tailin Jin, and 6 more authors

   2024

   Bib Paper

   @misc{huang2024resiliencemultiagentsystemsmalicious,
     title = {On the Resilience of Multi-Agent Systems with Malicious Agents},
     author = {Huang, Jen-tse and Zhou, Jiaxu and Jin, Tailin and Zhou, Xuhui and Chen, Zixi and Wang, Wenxuan and Yuan, Youliang and Sap, Maarten and Lyu, Michael R.},
     year = {2024},
     eprint = {2408.00989},
     archiveprefix = {arXiv},
     primaryclass = {cs.AI},
   }

5. NeurIPS

   Consent in Crisis: The Rapid Decline of the AI Data Commons

   In NeurIPS, 2024

   Bib Paper Website

   @inproceedings{consent2024,
     title = {Consent in Crisis: The Rapid Decline of the AI Data Commons},
     author = {},
     year = {2024},
     booktitle = {NeurIPS},
   }

6. COLM

   PolygloToxicityPrompts: Multilingual Evaluation of Neural Toxic Degeneration in Large Language Models

   Devansh Jain, Priyanshu Kumar, Samuel Gehman, and 3 more authors

   In COLM, 2024

   Paper
7. EMNLP

   Is this the real life? Is this just fantasy? The Misleading Success of Simulating Social Interactions With LLMs

   Xuhui Zhou, Zhe Su, Tiwalayo Eisape, and 2 more authors

   In EMNLP, 2024

   Bib Paper

   @inproceedings{zhou2024reallifejustfantasy,
     title = {Is this the real life? Is this just fantasy? The Misleading Success of Simulating Social Interactions With LLMs},
     author = {Zhou, Xuhui and Su, Zhe and Eisape, Tiwalayo and Kim, Hyunwoo and Sap, Maarten},
     year = {2024},
     eprint = {2403.05020},
     booktitle = {EMNLP},
     primaryclass = {cs.CL},
   }

8. ICLR

   Can LLMs Keep a Secret? Testing Privacy Implications of Language Models via Contextual Integrity Theory

   Niloofar Mireshghallah, Hyunwoo Kim, Xuhui Zhou, and 4 more authors

   In ICLR, 2024

   Bib Paper Website

   Spotlight (top 5%)

   @inproceedings{Niloofar2024CanLLMsKeepASecret,
     title = {Can LLMs Keep a Secret? Testing Privacy Implications of Language Models via Contextual Integrity Theory},
     author = {Mireshghallah, Niloofar and Kim, Hyunwoo and Zhou, Xuhui and Tsvetkov, Yulia and Sap, Maarten and Shokri, Reza and Choi, Yejin},
     booktitle = {ICLR},
     year = {2024},
     url = {https://openreview.net/forum?id=gmg7t8b4s0},
     award = {Spotlight (top 5%)}
   }

9. ICLR

   SOTOPIA: Interactive Evaluation for Social Intelligence in Language Agents

   Xuhui Zhou*, Hao Zhu*, Leena Mathur, and 8 more authors

   In ICLR, 2024

   Bib Paper Website

   Spotlight (top 5%)

   @inproceedings{zhou2024sotopia,
     title = {SOTOPIA: Interactive Evaluation for Social Intelligence in Language Agents},
     author = {Zhou*, Xuhui and Zhu*, Hao and Mathur, Leena and Zhang, Ruohong and Qi, Zhengyang and Yu, Haofei and Morency, Louis-Philippe and Bisk, Yonatan and Fried, Daniel and Neubig, Graham and Sap, Maarten},
     booktitle = {ICLR},
     year = {2024},
     url = {https://openreview.net/forum?id=mM7VurbA4r},
     award = {Spotlight (top 5%)},
   }

10. ICLR

    WebArena: A Realistic Web Environment for Building Autonomous Agents

    Shuyan Zhou, Frank F. Xu, Hao Zhu, and 8 more authors

    In ICLR, 2024

    Bib Paper Website Demo

    @inproceedings{zhou2024arena,
      title = {WebArena: A Realistic Web Environment for Building Autonomous Agents},
      author = {Zhou, Shuyan and Xu, Frank F. and Zhu, Hao and Zhou, Xuhui and Lo, Robert and Sridhar, Abishek and Cheng, Xianyi and Bisk, Yonatan and Fried, Daniel and Alon, Uri and Neubig, Graham},
      booktitle = {ICLR},
      year = {2024},
      url = {https://openreview.net/forum?id=oKn9c6ytLx},
      demo = {https://webarena.dev/#try-it-yourself}
    }

11. EACL

    Clever Hans or Neural Theory of Mind? Stress Testing Social Reasoning in Large Language Models

    Natalie Shapira, Mosh Levy, Hossein Seyed Alavi, and 5 more authors

    In EACL, 2024

    Bib Paper

    @inproceedings{shapira2024cleverHans,
      title = {Clever Hans or Neural Theory of Mind? Stress Testing Social Reasoning in Large Language Models},
      author = {Shapira, Natalie and Levy, Mosh and Seyed Alavi, Hossein and Zhou, Xuhui and Choi, Yejin and Goldberg, Yoav and Sap, Maarten and Shwartz, Vered},
      year = {2024},
      booktitle = {EACL},
    }

2023

1. ACL

   COBRA 🐍 Frames: Contextual Reasoning about Effects and Harms of Offensive Statements

   Xuhui Zhou, Hao Zhu, Akhila Yerukola, and 4 more authors

   In Findings of ACL, 2023

   Bib Paper Website Demo

   @inproceedings{zhou2023cobra,
     title = {COBRA 🐍 Frames: Contextual Reasoning about Effects and Harms of Offensive Statements},
     author = {Zhou, Xuhui and Zhu, Hao and Yerukola, Akhila and Davidson, Thomas and D. Hwang, Jena and Swayamdipta, Swabha and Sap, Maarten},
     year = {2023},
     booktitle = {Findings of ACL},
     demo = {https://cobra.allen.ai/},
   }

2. EMNLP

   “Don’t Take This Out of Context!” On the Need for Contextual Models and Evaluations for Stylistic Rewriting

   Akhila Yerukola, Xuhui Zhou, and Maarten Sap

   In EMNLP, 2023

   Bib Paper

   @inproceedings{yerukola2023contextRewrite,
     title = {``Don't Take This Out of Context!'' On the Need for Contextual Models and Evaluations for Stylistic Rewriting},
     author = {Yerukola, Akhila and Zhou, Xuhui and Sap, Maarten},
     year = {2023},
     booktitle = {EMNLP},
   }

3. EMNLP

   FANToM: A Benchmark for Stress-testing Machine Theory of Mind in Interactions

   Hyunwoo Kim, Melanie Sclar, Xuhui Zhou, and 4 more authors

   In EMNLP, 2023

   Bib Paper Website

   @inproceedings{kim2023fantom,
     title = {FANToM: A Benchmark for Stress-testing Machine Theory of Mind in Interactions},
     author = {Kim, Hyunwoo and Sclar, Melanie and Zhou, Xuhui and Le Bras, Ronan and Kim, Gunhee and Choi, Yejin and Sap, Maarten},
     year = {2023},
     booktitle = {EMNLP},
   }

4. EMNLP

   Learning to translate by learning to communicate

   C. Downey*, Xuhui Zhou*, L. Liu, and 1 more author

   In EMNLP MRL, 2023

   Bib Paper

   @inproceedings{Downey2022LearningTT,
     title = {Learning to translate by learning to communicate},
     author = {Downey*, C. and Zhou*, Xuhui and Liu, L. and Steinert-Threlkeld, Shane},
     booktitle = {EMNLP MRL},
     year = {2023}
   }

2022

1. NACCL

   Annotators with Attitudes: How Annotator Beliefs And Identities Bias Toxic Language Detection

   Maarten Sap, Swabha Swayamdipta, Laura Vianna, and 3 more authors

   In NAACL, 2022

   Bib Paper

   @inproceedings{sap2022annotatorsWithAttitudes,
     title = {Annotators with Attitudes: How Annotator Beliefs And Identities Bias Toxic Language Detection},
     author = {Sap, Maarten and Swayamdipta, Swabha and Vianna, Laura and Zhou, Xuhui and Choi, Yejin and Smith, Noah A.},
     year = {2022},
     booktitle = {NAACL},
     journal = {NAACL}
   }

2. ACL ConvAI

   Extracting and Inferring Personal Attributes from Dialogue

   Zhilin Wang, Xuhui Zhou, Rik Koncel-Kedziorski, and 2 more authors

   In ACL ConvAI, 2022

   Bib Paper

   @inproceedings{zhilin2022exandinfpersonalatt,
     title = {Extracting and Inferring Personal Attributes from Dialogue},
     author = {Wang, Zhilin and Zhou, Xuhui and Koncel-Kedziorski, Rik and Marin, Alex and Xia, Fei},
     year = {2022},
     booktitle = {ACL ConvAI}
   }

3. ICLR EmeCom

   Emergent Communication Fine-tuning (EC-FT) for Pretrained Language Models

   Shane Steinert-Threlkeld, Xuhui Zhou, Zeyu Liu, and 1 more author

   In Emergent Communication Workshop at ICLR 2022, 2022

   Bib Paper

   Runner-up Best Paper

   @inproceedings{steinert-threlkeld2022emergent,
     title = {Emergent Communication Fine-tuning ({EC}-{FT}) for Pretrained Language Models},
     author = {Steinert-Threlkeld, Shane and Zhou, Xuhui and Liu, Zeyu and Downey, C.M.},
     booktitle = {Emergent Communication Workshop at ICLR 2022},
     award = {Runner-up Best Paper},
     year = {2022},
     url = {https://openreview.net/forum?id=SUqrM7WR7W5}
   }

2021

1. EACL

   Challenges in Automated Debiasing for Toxic Language Detection

   Xuhui Zhou, Maarten Sap, Swabha Swayamdipta, and 2 more authors

   In EACL, 2021

   Bib Paper Code

   @inproceedings{zhou-etal-2020-debiasing,
     title = {Challenges in Automated Debiasing for Toxic Language Detection},
     author = {Zhou, Xuhui and Sap, Maarten and Swayamdipta, Swabha and Choi, Yejin and Smith, Noah A.},
     booktitle = {EACL},
     year = {2021},
   }

2020

1. BlackboxNLP

   Linguistically-Informed Transformations (LIT): A Method for Automatically Generating Contrast Sets

   Chuanrong Li, Lin Shengshuo, Zeyu Liu, and 3 more authors

   In Proceedings of the Third BlackboxNLP Workshop on Analyzing and Interpreting Neural Networks for NLP, Nov 2020

   Abs Paper

   Although large-scale pretrained language models, such as BERT and RoBERTa, have achieved superhuman performance on in-distribution test sets, their performance suffers on out-of-distribution test sets (e.g., on contrast sets). Building contrast sets often requires human-expert annotation, which is expensive and hard to create on a large scale. In this work, we propose a Linguistically-Informed Transformation (LIT) method to automatically generate contrast sets, which enables practitioners to explore linguistic phenomena of interests as well as compose different phenomena. Experimenting with our method on SNLI and MNLI shows that current pretrained language models, although being claimed to contain sufficient linguistic knowledge, struggle on our automatically generated contrast sets. Furthermore, we improve models’ performance on the contrast sets by applying LIT to augment the training data, without affecting performance on the original data.

2. EMNLP

   Multilevel Text Alignment with Cross-Document Attention

   Xuhui Zhou, Nikolaos Pappas, and Noah A. Smith

   In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), Nov 2020

   Abs Paper Website

   Text alignment finds application in tasks such as citation recommendation and plagiarism detection. Existing alignment methods operate at a single, predefined level and cannot learn to align texts at, for example, sentence \textitand document levels. We propose a new learning approach that equips previously established hierarchical attention encoders for representing documents with a cross-document attention component, enabling structural comparisons across different levels (document-to-document and sentence-to-document). Our component is weakly supervised from document pairs and can align at multiple levels. Our evaluation on predicting document-to-document relationships and sentence-to-document relationships on the tasks of citation recommendation and plagiarism detection shows that our approach outperforms previously established hierarchical, attention encoders based on recurrent and transformer contextualization that are unaware of structural correspondence between documents.

3. ACL SRW

   RPD: A Distance Function Between Word Embeddings

   Xuhui Zhou, Shujian Huang, and Zaixiang Zheng

   In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics: Student Research Workshop, Nov 2020

   Abs Paper

   It is well-understood that different algorithms, training processes, and corpora produce different word embeddings. However, less is known about the relation between different embedding spaces, i.e. how far different sets of em-beddings deviate from each other. In this paper, we propose a novel metric called Relative Pairwise Inner Product Distance (RPD) to quantify the distance between different sets of word embeddings. This unitary-invariant metric has a unified scale for comparing different sets of word embeddings. Based on the properties of RPD, we study the relations of word embeddings of different algorithms systematically and investigate the influence of different training processes and corpora. The results shed light on the poorly understood word embeddings and justify RPD as a measure of the distance of embedding space.

4. AAAI

   Evaluating Commonsense in Pre-trained Language Models

   Xuhui Zhou, Y. Zhang, Leyang Cui, and 1 more author

   In Proceedings of the AAAI Conference on Artificial Intelligence, 34(05), Nov 2020

   Paper