Search

Peter Henderson
  • Home
  • Talks
  • Publications
  • Press
Avatar

Peter Henderson

JD/PhD Candidate

Stanford University

About Me

I’m a joint JD-PhD (Computer Science) candidate at Stanford University where I’m lucky enough to be advised by Dan Jurafsky. I’m also an OpenPhilanthropy AI Fellow and a Graduate Student Fellow at the Regulation, Evaluation, and Governance Lab. At Stanford Law School, I co-led the Domestic Violence Pro Bono Project, worked on client representation with the Three Strikes Project, and contributed to the Stanford Native Law Pro Bono Project. Previously, I was lucky enough to be advised by David Meger and Joelle Pineau for my M.Sc. at McGill University and the Montréal Institute for Learning Algorithms.

I also spent time as a Software Engineer and Applied Scientist at Amazon AWS/Alexa, worked with Justice Cuéllar at the California Supreme Court, am a part-time researcher with the Internal Revenue Service Research, Applied Analytics and Statistics Division, and am a Technical Advisor at the Institute for Security+Technology.

My research focuses on the intersection of AI and Law, in particular creating robust decision-making systems and regulatory frameworks for safe ML deployments. Some of my work has received coverage by TechCrunch, Science, The Wall Street Journal, Bloomberg, and more. I also occassionally post a roundup of all thing AI policy, the latest of which is here. Overall, I’m interested in a wide range of other technical machine learning research, policy, and legal work, so get in touch if you’d like to collaborate!

Latest News

  • Integrating Reward Maximization and Population Estimation: Sequential Decision-Making for Internal Revenue Service Audit Selection has been accepted to AAAI 2023! Workshop versions previously appeared at the Adaptive Experimental Design and Active Learning in the Real World Workshop @ ICML 2022 and the 12th Annual IRS/TPC Joint Research Conference on Tax Administration. It’s an example of RL (okay, well bandits) used in the public sector in a real world setting, taking into account real policy objectives in an under-explored setting of algorithms. You can see my talk on it here and some coverage by Reuters here.
  • Our concurrent paper Entropy Regularization for Population Estimation also draws out the connections of the optimize-and-estimate structured bandit setting to entropy regularization, also to appear at AAAI 2023!
  • Holistic Evaluation of Language Models is now online. You can check out: the accompanying webpage, blog post, as well as coverage by VentureBeat.
  • Pile of Law: Learning Responsible Data Filtering from the Law and a 256GB Open-Source Legal Dataset is out! Check it out, as well as the accompanying HAI Blog Post. It has been featured in Bloomberg, FiveThirtyEight and accepted as an oral presentation to NeurIPS Datasets & Benchmarks Track!
  • My recent panel discussion on AI and sustainability received some coverage at TechTarget and VentureBeat! You can see the video here.
  • Vulnerabilities in Discovery Tech is now officially in the Harvard Journal of Law & Technology (2022)! We investigate how ML is used in discovery proceedings, identify potential risks and vulnerabilities, and suggest a path forward for ensuring robust use of ML in civil litigation. It has been noted in Jotwell as “required reading for anyone who hopes that artificial intelligence will harness and solve the problems of discovery in document-intensive litigation.”

Peter Henderson © 2022 · Powered by the Academic theme for Hugo.

Cite
Copy Download