--- title: "Pang Wei Koh — author profile and papers" description: "Pang Wei Koh, University of Wisconsin–Madison: 2 papers on Elamur Research (2017–2023), h-index 28 — affiliations, topics, co-authors and who cites them…" canonical: "https://research.elamur.ai/free/en/authors/pang-wei-koh/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/authors/pang-wei-koh/) [home](https://md.research.elamur.ai/free/en/) / [authors](https://md.research.elamur.ai/free/en/authors/) / Pang Wei Koh # Pang Wei Koh Pang Wei Koh works on understanding and evaluating machine learning models. Their papers here take up two recurring questions: where a model's predictions come from, and how far its outputs can be trusted. The methods range from statistical tools that trace a prediction back to the training set to evaluation schemes that check generated text against an external knowledge source. The later work falls under [text generation](https://md.research.elamur.ai/free/en/topics/text-generation/) with large language models. The most cited of these papers is [Understanding Black-box Predictions via Influence Functions](https://md.research.elamur.ai/free/en/papers/koh-2017-understanding-black-box-predictions-influence/) (2017), written with [Percy Liang](https://md.research.elamur.ai/free/en/authors/percy-liang/). Most explanation methods ask how a trained model arrives at an output. This paper asks instead which training examples were responsible for it. It adapts influence functions from robust statistics to estimate how the parameters and the loss on a test point would shift if one training example were up-weighted, removed or perturbed. It avoids retraining the model and never forms or inverts the Hessian explicitly, which makes the approach workable for models with many parameters. The second paper, [FActScore](https://md.research.elamur.ai/free/en/papers/min-2023-factscore-fine-grained-atomic-evaluation/) (2023), deals with factual accuracy in long generated passages. It breaks a response into atomic facts, each carrying a single piece of information, and reports the share of them that a chosen knowledge source supports. The paper tests this on biographies of people, using Wikipedia as the reference, and compares several language models using expert annotations. The move from the 2017 paper to the 2023 one shifts the focus from explaining individual predictions to measuring how reliable large language models are. Koh's earlier work was done at [Stanford University](https://md.research.elamur.ai/free/en/universities/stanford-university/), and the current affiliations are the [University of Washington](https://md.research.elamur.ai/free/en/universities/university-of-washington/) and the [Allen Institute for Artificial Intelligence](https://md.research.elamur.ai/free/en/labs/allen-institute-for-artificial-intelligence/) in Seattle. Co-authors on the factuality work include [Sewon Min](https://md.research.elamur.ai/free/en/authors/sewon-min/), [Luke Zettlemoyer](https://md.research.elamur.ai/free/en/authors/luke-zettlemoyer/) and [Hannaneh Hajishirzi](https://md.research.elamur.ai/free/en/authors/hannaneh-hajishirzi/). Pang Wei Koh has 2 papers in our library, in 2017–2023, most of them on [Text generation](https://md.research.elamur.ai/free/en/topics/text-generation/); often with [Hannaneh Hajishirzi](https://md.research.elamur.ai/free/en/authors/hannaneh-hajishirzi/), [Kalpesh Krishna](https://md.research.elamur.ai/free/en/authors/kalpesh-krishna/), [Luke Zettlemoyer](https://md.research.elamur.ai/free/en/authors/luke-zettlemoyer/). 2 papers in our library · 2017–2023 [OpenAlex](https://openalex.org/A5079904764) · [ORCID](https://orcid.org/0000-0003-4330-6969) · [**h-index 28**](https://openalex.org/authors/A5079904764) · [110 works](https://openalex.org/works?filter=author.id:A5079904764) · [9,056 citations](https://openalex.org/works?filter=author.id:A5079904764&sort=cited_by_count:desc) OpenAlex, 2026-10-14 [📄 2 works in our library](https://md.research.elamur.ai/free/en/authors/pang-wei-koh/#papers) · [🔗 2,552 citations in our library](https://md.research.elamur.ai/free/en/authors/pang-wei-koh/cited-by/) · [10 more in our library without a page yet](#also) · [⭐ Paper Star rank #1008](https://md.research.elamur.ai/free/en/paper-stars/page/11/#star-1008) The profile is the library's match of the name; two people with one name are not always told apart. Elsewhere: [Google Scholar](https://scholar.google.com/citations?view_op=search_authors&mauthors=Pang+Wei+Koh) · [Semantic Scholar](https://www.semanticscholar.org/search?q=Pang+Wei+Koh) · [dblp](https://dblp.org/search?q=Pang+Wei+Koh) · [ACL Anthology](https://aclanthology.org/search/?q=Pang+Wei+Koh) ## Affiliations [🎓**University of Washington** Seattle · United States 2023–2026 · 31 papers*current*](https://md.research.elamur.ai/free/en/universities/university-of-washington/) · [🔬**Allen Institute for Artificial Intelligence** Seattle · United States 2024–2026 · 14 papers*current*](https://md.research.elamur.ai/free/en/labs/allen-institute-for-artificial-intelligence/) · [🎓**University of Wisconsin–Madison** Madison · United States 2026 · 1 paper*current*](https://md.research.elamur.ai/free/en/universities/university-of-wisconsin-madison/) · [🎓**Stanford University** Stanford · United States 2010–2021 · 25 papers](https://md.research.elamur.ai/free/en/universities/stanford-university/) · [🏢**Google DeepMind** London · United Kingdom 2023–2024 · 2 papers](https://md.research.elamur.ai/free/en/companies/google-deepmind/) · [🏥**Stanford Medicine** Stanford · United States 2016–2017 · 2 papers](https://md.research.elamur.ai/free/en/hospitals/stanford-medicine/) From the bylines of their papers (OpenAlex, 2026-10-14), ORCID and Wikidata; institutions on papers are matched by machine. 1 more institution on single bylines are not listed. ## Topics [Text generation (1)](https://md.research.elamur.ai/free/en/topics/text-generation/) ## Works by year ### 2020s [**2023** 1 paper](https://md.research.elamur.ai/free/en/papers/min-2023-factscore-fine-grained-atomic-evaluation/) ### 2010s [**2017** 1 paper](https://md.research.elamur.ai/free/en/papers/koh-2017-understanding-black-box-predictions-influence/) ## Papers The most cited first (OpenAlex and our library). 1. **[Understanding Black-box Predictions via Influence Functions](https://md.research.elamur.ai/free/en/papers/koh-2017-understanding-black-box-predictions-influence/)** *(2017 · 1,478 citations)* Explanation methods for black-box models usually ask how a fixed, already-trained model reaches a particular output. 2. **[FActScore: Fine-grained Atomic Evaluation of Factual Precision in Long Form Text Generation](https://md.research.elamur.ai/free/en/papers/min-2023-factscore-fine-grained-atomic-evaluation/)** *(2023 · 970 citations)* Checking the factuality of long outputs from large language models is hard for two reasons. One response usually mixes correct and incorrect claims, so a single yes/no verdict per sentence or response discards useful information. Verifying every claim by hand is also slow and expensive. The authors propose FActScore… ## Also in our library 10 more works of theirs are in our library without a page here yet; they will appear as we publish them. ## Co-authors here 9 co-authors with a page, the most papers together first. - [Hannaneh Hajishirzi](https://md.research.elamur.ai/free/en/authors/hannaneh-hajishirzi/) - [Kalpesh Krishna](https://md.research.elamur.ai/free/en/authors/kalpesh-krishna/) - [Luke Zettlemoyer](https://md.research.elamur.ai/free/en/authors/luke-zettlemoyer/) - [Mike Lewis](https://md.research.elamur.ai/free/en/authors/mike-lewis/) - [Mohit Iyyer](https://md.research.elamur.ai/free/en/authors/mohit-iyyer/) - [Percy Liang](https://md.research.elamur.ai/free/en/authors/percy-liang/) - [Sewon Min](https://md.research.elamur.ai/free/en/authors/sewon-min/) - [Wen-tau Yih](https://md.research.elamur.ai/free/en/authors/wen-tau-yih/) - [Xinxi Lyu](https://md.research.elamur.ai/free/en/authors/xinxi-lyu/) ## Cited by Papers of our library cite their works 2,552 times (of the 140,884 papers whose bibliographies we read); [198 citing papers with a page here →](https://md.research.elamur.ai/free/en/authors/pang-wei-koh/cited-by/).