Congratulations to Dr. Abhijnan Nath!
26 February 2026
Congratulations to Dr. Abhijnan Nath for successfully defending his Ph.D. dissertation!
Abhijnan’s dissertation, The Generative Friction Trilogy: Learning to Intervene, Evaluate, and Collaborate in Multi-Agent Collaboration, posits that aligning modern language models for collaborative tasks requires treating “friction” as a productive and necessary component of theoretical modeling, training, and evaluation. “Friction” in human-AI interaction has recently entered the mainstream lexicon as a way to decrease overreliance and make humans think for themselves during their interactions with AI systems. Abhijnan’s work was among the earliest to approach this from the modeling and training perspective, backed up with deep theory and proofs, as well as compelling empirical results. The bulk of his dissertation has been published as papers at top conferences including ACL, NeurIPS, and AAMAS, and has been cited in Microsoft’s 2025 New Future of Work report as an example of research that can help create a better future of work with AI.
Congratulations to Abhijnan! Your work is ahead of the curve and stands to make a great impact on the field. It’s already made a huge impact on the lab. The bar is high. Best of luck in what comes next!
Here is a picture after defense, of Abhijnan with two of his committee members: Prof. Ewan Davies (CS) and Prof. Michael Kirby (Math). And me (not shown, Prof. Nathaniel Blanchard, who attended online).
In the words of Ewan: “I want to give a Ph.D. to this man.”
(X-posted on signallab.ai)