Tanya Goyal

tanyagoyal@cornell.edu

I am an assistant professor in the Department of Computer Science at Cornell University. Previously, I was a postdoctoral scholar at Princeton Language and Intelligence Center (2023-2024). I obtained my Ph.D. in Computer Science at UT Austin in 2023, advised by Greg Durrett. My Ph.D. research focused on building evaluation tools for text generation models, and was awarded UTCS’s Bert Kay Dissertation award. Before that, I learnt how to Math at the Indian Institute of Technology, Guwahati (Maths and Computing, B.Tech 2011-2015).

My research interests are primarily in the field of Natural Language Processing. Problems that I am currently excited about include:

  1. Reliable and sustainable evaluation frameworks for large language models (LLMs).
  1. Factuality assessment and improvement of LLMs.
  1. Understanding LLM behaviors as a function of training data and/or aligment strategies.

See more at 📚 Publications.

Note to prospective students: Thank you for your interest in our group’s research! If you are a Cornell student, please email me. If not, see Cornell CS’s PhD admission page for details of the admission process. I am looking to admit multiple PhD students in the next admissions cycle.

Recent News
  • 2024/07 - Started at Cornell, moved to Ithaca!
  • 2024/06 - Wrapped up postdoc at Princeton.
  • 2024/06 - Gave a talk at Weill Cornell Medicine CTSC lecture series for the Introduction to Generative AI/LLM.
  • 2024/05 - Attended ICLR in Vienna

Publications

  • Challenges in Trustworthy Human Evaluation of Chatbots [PDF]
    Wenting Zhao, Sasha Rush, Tanya Goyal.
    NAACL Findings 2025.

  • (arxiv) Recycled Attention: Efficient Infrence for Long-Context Language Models [PDF]
    Fangyuan Xu, Tanya Goyal, Eunsol Choi.
    arxiv 2024.

  • (arxiv) WildHallucinations: Evaluating Long-form Factuality in LLMs with Real-World Entity Queries [PDF]
    Wenting Zhao, Tanya Goyal, Yu Ying Chiu, Liwei Jiang, Benjamin Newman, Abhilasha Ravichander, Khyathi Chandu, Ronan Le Bras, Claire Cardie, Yuntian Deng, Yejin Choi.
    arxiv 2024.

  • One Thousand and One Pairs: A "novel" challenge for long-context language models [PDF] [Leaderboard]
    Marzena Karpinska, Katherine Thai, Kyle Lo, Tanya Goyal, Mohit Iyyer.
    EMNLP 2024.

  • LitSearch: A Retrieval Benchmark for Scientific Literature Search [PDF]
    Anirudh Ajith, Mengzhou Xia, Alexis Chevalier, Tanya Goyal, Danqi Chen, Tianyu Gao.
    EMNLP 2024.

  • D2PO: Discriminator-Guided DPO with Response Evaluation Models [PDF]
    Prasann Singhal, Nathan Lambert, Scott Niekum, Tanya Goyal, and Greg Durrett.
    COLM 2024.
  • A Long Way To Go: Investigating Length Correlations in RLHF [PDF]
    Prasann Singhal, Tanya Goyal, Jiacheng Xu, Greg Durrett.
    COLM 2024.

  • FABLES: Evaluating faithfulness and content selection in book-length summarization [PDF]
    Yekyung Kim, Yapei Chang, Marzena Karpinska, Aparna Garimella, Varun Manjunatha, Kyle Lo, Tanya Goyal, Mohit Iyyer.
    COLM 2024.

  • BooookScore: A systematic exploration of book-length summarization in the era of LLMs [PDF] [Code]
    Yapei Chang, Kyle Lo, Tanya Goyal, Mohit Iyyer.
    ICLR 2024.

  • Evaluating Large Language Models at Evaluating Instruction Following [PDF]
    Zhiyuan Zeng, Jiatong Yu, Tianyu Gao, Yu Meng, Tanya Goyal, Danqi Chen.
    ICLR 2024.

  • Fine-grained evaluation for text summarization [PDF]
    Tanya Goyal.
    Thesis, University of Texas at Austin, 2023.
    (awarded UTCS’s Bert Kay Dissertation Award)

  • WiCE: Real-World Entailment for Claims in Wikipedia [PDF]
    Ryo Kamoi, Tanya Goyal, Juan Diego Rodriguez, and Greg Durrett.
    EMNLP 2023.

  • (arxiv) News Summarization and Evaluation in the Era of GPT-3 [PDF]
    Tanya Goyal, Junyi Jessy Li, Greg Durrett.
    arxiv 2022.

  • Understanding Factual Errors in Summarization- Errors, Summarizers, Datasets, Error Detectors [PDF]
    Liyan Tang, Tanya Goyal, Alexander R. Fabbri, Philippe Laban, Jiacheng Xu, Semih Yahvuz, Wojciech Kryściński, Justin F. Rousseau, and Greg Durrett.
    ACL 2023.

  • Shortcomings of Question Answering based Factuality Frameworks for Error Localization [PDF]
    Ryo Kamoi, Tanya Goyal, Greg Durrett.
    EACL 2023.

  • SNaC - Coherence Error Detection for Narrative Summarization [PDF]
    Tanya Goyal, Junyi Jessy Li, Greg Durrett.
    EMNLP 2022.

  • HydraSum: Disentangling Stylistic Features in Text Summarization using Multi-Decoder Models [PDF]
    Tanya Goyal, Nazneen Fatema Rajani, Wenhao Liu, Wojciech Kryściński.
    EMNLP 2022.

  • FALTE: A Toolkit for Fine-grained Annotation for Long Text Evaluation [PDF] [Code]
    Tanya Goyal, Junyi Jessy Li, Greg Durrett.
    EMNLP-Demo 2022.

  • Training Dynamics for Text Summarization Models [PDF]
    Tanya Goyal, Jiacheng Xu, Junyi Jessy Li, Greg Durrett.
    ACL-Findings 2022.

  • Annotating and Modeling Fine-grained Factuality in Summarization [PDF] [Slides] [Poster]
    Tanya Goyal, Greg Durrett.
    NAACL 2021.

  • Evaluating Factuality in Generation with Dependency-level Entailment [PDF] [Code] [Poster]
    Tanya Goyal, Greg Durrett.
    EMNLP-Findings 2020.

  • Neural Syntactic Preordering for Controlled Paraphrase Generation [PDF] [Code] [Slides]
    Tanya Goyal, Greg Durrett.
    ACL 2020.

  • Embedding Time Expressions in Deep Temporal Ordering Models [PDF]
    Tanya Goyal, Greg Durrett.
    ACL 2019.

  • Your Behavior Signals Your Reliability: Modeling Crowd Behavioral Traces to Ensure Quality Relevance Annotation [PDF]
    Tanya Goyal, Tyler McDonnell, Mucahid Kutlu, Tamer Elsayed, Matthew Lease.
    HCOMP 2018.

  • An Empirical Analysis of Edit Importance between Document Versions [PDF]
    Tanya Goyal, Sachin Kelkar, Manas Agarwal, Jeenu Grover.
    EMNLP 2017.

Teaching

📝CS 6740: Advanced Language Technologies (Fall 2024; graduate) 📝CS 4740: Natural Language Processing (Spring 2025; undergrad)