Steven Dillmann

PhD student at Stanford University working on AI for Science, affiliated with the Stanford AI Lab and ICME. His research focuses on foundation models, AI agents for scientific discovery, and rigorous evaluation of LLMs and agents on scientific tasks, including computational workflows in the natural sciences.

AI-for-Science Advances Depend on Evaluation, Not Just Generation

In a Stanford AI+Science lightning-talk session introduced by Surya Ganguli, four young researchers made a common case: AI-for-science is useful only when paired with rigorous evaluation. Aishwarya Mandyam, Amar Venugopal, Steven Dillmann and Alda Elfarsdóttir each treated AI systems or outputs as claims to be tested — through uncertainty estimates for clinical policies, causal checks on generated text, executable benchmarks for scientific agents, and empirical links between corporate climate language and later emissions.

Stanford HAIMay 15, 20267 min read