October Buzz: Papers from Our Scholars

Big congratulations to our researchers and faculty! 

Several of our scholars have recently published or released exciting new work spanning topics from glaciology to protein modeling, experimental design, and large language models. These papers—appearing in top journals and conferences, including PLOS Computational Biology, Wildlife Research, and the European Geosciences Union, as well as at the Conference on Digital Experimentation @ MIT (CODE@MIT)—showcase the incredible range and depth of data science research happening across our community.

David Bruns-Smith

• Two-Stage Machine Learning for Nonparametric Instrumental Variable Regression

• Augmented Balancing Weights as Linear Regression, Journal of the Royal Statistical Society Series B: Statistical Methodology, 2025. Presented as a keynote at the RSS Annual Conference in Edinburgh (September 2025).

Alexandra DiGiacomo

• Non-invasive extraction of white shark swimming kinematics from unoccupied aircraft system (UAS) imagery — Wildlife Research 

Luc Houriez 

• Reinforced ridges in Thwaites Glacier yield insights into resolution requirements for coupled ice sheet and solid Earth models — European Geosciences Union

Anuska Murthy

• When Does Interference Matter? Decision-Making in Platform Experiments

Big congratulations to Anuska, Ramesh Johari, and their co-authors for their paper acceptance at the Conference on Digital Experimentation @ MIT (CODE@MIT)! Ramesh is also giving a plenary talk at CODE@MIT, where he'll talk about this paper as well as some other related work he's done in this area.

Gowri Nayar

• Paying attention to attention: High attention sites as indicators of protein family and function in language models — PLOS Computational Biology 

Rylan Shaeffer

• Evaluating the Robustness of Chinchilla Compute-Optimal Scaling

• Pretraining Scaling Laws for Generative Evaluations of Language Models

• Understanding Adversarial Transfer: Why Representation-Space Attacks Fail Where Data-Space Attacks Succeed

Mahlet Shiferaw

• How Do Uncertainties in Galaxy Formation Physics Impact Field-level Galaxy Bias

Min Sun

• No Tokens Wasted: Leveraging Long Context in Biomedical Vision-Language Models

This article is a part of the Stanford Data Science legacy publication. Read more about the HAI and Stanford Data Science merger.