Vedang Lad
I'm an AI researcher — a blend of engineer and scientist. My interests span computer science, neuroscience, and math, with a particular focus on the mechanisms and structures that facilitate intelligence, and how we can leverage that understanding to build smarter, safer systems.
Currently, I'm a Founding ML Researcher at Metamorphic, a startup in Palo Alto that spun out of Stanford, where I help build multimodal world models. Before that, I was a MATS scholar and studied computer science and physics at MIT. There, I worked on interpretability with Max Tegmark and spent time on data-centric methods in ML. I'll be starting a PhD at Stanford in Fall 2026.
Outside of research, I was an NCAA track and field athlete at MIT and still run competitively for Peninsula Distance Club in the Bay Area.
Currently, I'm a Founding ML Researcher at Metamorphic, a startup in Palo Alto that spun out of Stanford, where I help build multimodal world models. Before that, I was a MATS scholar and studied computer science and physics at MIT. There, I worked on interpretability with Max Tegmark and spent time on data-centric methods in ML. I'll be starting a PhD at Stanford in Fall 2026.
Outside of research, I was an NCAA track and field athlete at MIT and still run competitively for Peninsula Distance Club in the Bay Area.
Papers
Letting the Neural Code Speak: Automated Characterization of Monkey Visual Neurons Through Human Language
Vedang Lad, Katrin Franke, Tamar Rott Shaham, Surya Ganguli, Andreas S. Tolias, Sophia Sanborn, Nikos Karantzas
The Remarkable Robustness of LLMs: Stages of Inference?
Vedang Lad, Jin Hwa Lee, Wes Gurnee, Max Tegmark
Exploring the Integration of AI into Physics Education: Leveraging ChatGPT for Problem Generation
Vedang Lad, Isaac Liao, Mohamed Abdelhafez, Peter Dourmashkin, Saif El-Adawy
Opening the AI Black Box: Distilling Machine-Learned Algorithms into Code
Eric J. Michaud, Isaac Liao, Vedang Lad, Ziming Liu, Anish Mudide, Chloe Loughridge, Zifan Carl Guo, Tara Rezaei Kheirkhah, Mateja Vukelić, Max Tegmark
Estimating label quality and errors in semantic segmentation data via any model
Vedang Lad, Jonas Mueller