Join us in congratulating Dartmouth CS Professor Peter Winkler on being named 2022 AAAS Fellow! 🎉🎉 faculty.dartmouth.edu/artsandscience…

Great article in the Dartmouth News about the conversation around ChatGPT Dartmouth 🌲. home.dartmouth.edu/news/2023/02/c…

Thanks to the cohosts Roberto Rey Agudo, Tania Convertini, and TheLeslieCenter for moderating and hosting this panel discussion last Friday. The session proved intellectually stimulating, as diverse viewpoints were shared on ChatGPT and its impact on Language Education.

Training LLMs via Simulated Human Societies -Fine-grained social interaction data collected by running open-source simulated society platform -Collective ratings, detailed feedback, & revised responses fine-tune LLM -Reduces instability & reward gaming arxiv.org/abs/2305.16960

Excellent work by my Ph.D. student, Ruibo Liu (Ruibo Liu). His work on Stable Alignment nicely integrates the various areas he has been investigating throughout his Ph.D. journey and contributes significantly to making Language Models more prosocial - a key focus in our lab.

I strongly encourage everyone to experiment with the available code and integrate it into your research. Ruibo and I are happy to answer any questions you may have about the work.

Great piece by Neil Savage in Communications of the ACM on how chatbots can be used for malicious purposes and how technology may help detect #chatbot related content. Thrilled to be featured in it, along with the amazing Soroush Vosoughi, Tom Goldstein, @bviswana

Also, check out our paper in ACL, a great collaboration with Soroush Vosoughi and the team! aclanthology.org/2023.acl-short…

Thrilled to join Dartmouth's Center for Precision Health and Artificial Intelligence (CPHAI) as a Technical Associate Director, working under the leadership of Prof. Saeed Hassanpour (Saeed Hassanpour). #PrecisionHealth #AI #Dartmouth

Recently got quoted in an insightful MIT Technology Review article exploring the political bias in language models. A must-read for anyone interested in the intersection of tech and society. #AI #LLM #TechEthics #DartmouthCS technologyreview.com/2023/08/07/107…

In line with the MIT Technology Review article, don't miss the The Washington Post piece on ChatGPT's political leanings. It discusses our best-paper-winning study on political bias in language models & ways to mitigate this bias, published in #AAAI 2021. #DartmouthCS washingtonpost.com/technology/202…

Defining Participation Bias in Social Media - Dartmouth - ... a computer science postdoc and visiting scholar at Dartmouth, and Soroush Vosoughi, assistant professor of computer science. Unlike surveys, which ... - ift.tt/EmjTAH9

Happy to see our lab's work featured in a nature article on ethical AI by Neil Savage (Neil Savage): "How Robots Can Learn to Follow a Moral Code."

Delighted to see our latest research on "Deciphering Stereotypes in Pre-Trained Language Models," featured in Dartmouth News. Our study unveils key insights into how PLMs encode biases and proposes methods for their mitigation. #NLP #MachineLearning #ACL2023

Congratulations to my students Yiren Jian (Yiren Jian) and Weicheng Ma on the recent successful defense of their respective Ph.D. theses!

Excited to have 6 papers accepted to EMNLP 2024 (3 in the main conference, 3 in Findings), covering topics such as the working memory limitations of LLMs and introducing a new task and dataset for studying intellectual humility in online discourse. Congrats to all coauthors!

Very interesting work by MMS student Ivory Yang on using AI to revitalize endangered languages.