Daniel Khalil

I'm an undergraduate at the California Institute of Technology (Caltech) and a Student Researcher at Google Research.

At Caltech I have worked in the Vision Lab, and I currently work with Yisong Yue and Frances Arnold. I am also a researcher in the Berkeley AI Lab (BAIR).

Email  /  CV  /  Bio  /  Scholar  /  Twitter  /  Github

profile photo

Research

I'm interested in creating general solutions to impactful problems with machine learning. Specifically, I am interested in reinforcement learning, continual/representation learning, and other stuff. Representative papers are highlighted.

Steering Generative Models with Experimental Data for Protein Fitness Optimization
Jason Yang, Wenda Chu, Daniel Khalil, Raul Astudillo, Bruce J. Wittmann, Frances H. Arnold, Yisong Yue
NeurIPS, 2025 & ICLR GEMBIO Workshop, 2025
paper / code

Learning Keypoints for Multi-Agent Behavior Analysis using Self-Supervision
Daniel Khalil, Christina Liu, Pietro Perona, Jennifer Sun, Markus Marks
WACV, 2025   (Oral Presentation)
project page / video / arXiv

Projects

Here are some projects. Representative papers are highlighted.

Chest X-Ray classifier
code

PyTorch implementations of several different classifiers for chest X-ray images, based on the multi-class and multi-label CheXpert dataset.

Miscellanea

caltech TA, IDS 157 (Statistical Inference) Spring 2025
TA, CS 159 (Advanced Topics in Machine Learning) Spring 2025
TA, CS38 (Algorithms) Spring 2024
TA, CS156a (Machine Learning) Fall 2024

Website from John Barron.