Neel Guha
PhD, Computer Science, Stanford University
JD, Stanford Law School
nguha [AT] cs.stanford.edu
I am a fifth year JD-PhD student in Computer Science at Stanford University (advised by Chris Ré). I'm a part of the Hazy Research Lab, Stanford Center for Research on Foundation Models, and RegLab. I graduated with a MS in Machine Learning from Carnegie Mellon University ('19) and a BS (with Honors) in Computer Science from Stanford University ('18). I am grateful to be supported by the Stanford Interdisciplinary Graduate Fellowship (SIGF) and the HAI Graduate Fellowship.
My research lies at the intersection of artificial intelligence/machine learning (AI/ML) and law. My work explores four questions:
- How can ML advance our understanding of the law and legal institutions? ML provides a novel way to generate empirical insights into legal questions using large-scale datasets. In my work, I've developed methods to classify state statutes creating private rights of action, and used these methods to test theories of private enforcement at the state level. In ongoing work, I am extending these methods to identity other types of state statutes, such as officer immunities and state laws creating foreign policy.
- How should we govern AI? I examine how AI systems, particularly in sensitive domains like healthcare, can be effectively regulated. My research includes frameworks for assessing liability for medical AI and analyzing the technical and institutional trade-offs associated with traditional regulatory interventions, such as disclosure and licensing.
- What types of legal reasoning can large language models (LLMs) perform? The answer to this question has implications for how lawyers and other legal institutions (e.g., agencies, courts) think about AI adoption. I’ve contributed to the development of benchmarks that evaluate LLM performance on diverse legal reasoning and data tasks. These include LegalBench, CaseHOLD, and LoCo.
- How can LLM performance be improved without labeled data or human supervision? I investigate techniques that leverage auxiliary information (e.g., embeddings) to enhance model performance in unsupervised or weakly supervised settings. This is particulary relevant for legal applications, where access to labeled data is challenging. Examples of this work include Smoothie, Embroid, and Bootleg.
Recent News
December 2024: AI Regulation Has Its Own Alignment Problem officially out in the George Washington Law Review! November 2024: Diego Zambrano and I have a short piece out in the Wisconsin Law Review talking about a new empirical project that uses LLMs to build an annotated database of state statutes! October 2024: Michelle Mello and I appeared on the Stanford Legal Podcast (hosted by Professor Pam Karlan and Professor Rich Ford) to talk about our work on AI liability in healthcare! September 2024: Our work on learning unsupervised routers for LLMs accepted to NeurIPS 2024! July 2024: Excited to contribute to Open Problems in Technical AI Governance! May 2024: Excited to contribute a chapter on benchmarking language models for legal applications to The Oxford Handbook on the Foundations and Regulation of Generative AI (OUP, 2024). May 2024: Two papers accepted to ICML 2024: (1) Prospectors, and (2) Long-Context Retrievers. January 2024: Understanding Liability Risk from Using Health Care Artificial Intelligence Tools out in The New England Journal of Medicine (with Michelle Mello). January 2024: Private Enforcement in the States out in University of Pennsylvania Law Review (with Diego Zambrano, Austin Peters, and Jeffrey Xia).Website design and template by Martin Saveski.