PeopleofSuperintelligence

Researcher

Farzan Farnia

Farzan Farnia is an AI researcher whose work spans minimax supervised learning, GAN theory, and evaluation of generative models. He received a Stanford Graduate Fellowship from 2013 to 2016 and won a 2014 prize for Stanford Electrical Engineering Ph.D. qualifying-exam performance; he holds degrees from Sharif University of Technology and Stanford. In 2016, he and David Tse introduced the minimax SVM, and Farnia's later GAN research examined training theory and game equilibria. In 2026, he coauthored a method for measuring diversity in prompt-based generative models, and he is an assistant professor at CUHK as of October 2026.

Timeline

  1. 2013–2016

    A Stanford fellowship and prize

    He receives a Stanford Graduate Fellowship from 2013 to 2016. In 2014, he wins the Numerical Technology Founders Prize as the second top performer of Stanford Electrical Engineering Ph.D. qualifying exams.

  2. 2016

    The minimax SVM

    In 2016, he and David Tse introduce a minimax framework for supervised learning and derive the minimax SVM, which minimizes worst-case expected 0–1 loss.

  3. 2018

    A convex duality framework for GANs

    At NeurIPS 2018, he and David Tse develop a convex duality framework to analyze how constrained discriminators affect GAN training.

  4. 2019

    Electrical engineering at Sharif and Stanford

    He holds a B.Sc. in electrical engineering and mathematics from Sharif University of Technology and M.Sc. and Ph.D. degrees in electrical engineering from Stanford. At his May 2019 Ph.D. defense, he presents an information-theoretic framework for supervised learning.

  5. 2019–2021

    Postdoctoral work at MIT

    From 2019 to 2021, he works as a postdoctoral research associate at MIT's Laboratory for Information and Decision Systems.

  6. 2020

    Proximal training for GANs

    In 2020, he and Asuman Ozdaglar show that GAN zero-sum games may lack Nash equilibria and propose proximal equilibrium and proximal training.

  7. May 2026

    Prompt-aware diversity scoring

    In May 2026, he and Mohammad Jalali, Azim Ospanov, and Amin Gohari publish Conditional Vendi Score, a method for measuring prompt-induced and model-induced diversity in generative outputs.

  8. As of October 2026

    Assistant professor at CUHK

    As of October 2026, he is an assistant professor in CUHK's Department of Computer Science and Engineering; his research covers machine learning, deep learning theory, optimization, and information theory.

Career

  1. NowChinese University of Hong KongAssistant Professor
  2. 2019–2021Massachusetts Institute of TechnologyPostdoctoral Research Associate

Education

  1. 2019Stanford UniversityM.Sc.
  2. Sharif University of TechnologyB.Sc.

From the biography.

  1. The Massachusetts Institute of Technology trained or employed 44 of the 187 AI major-prize winners, more than any other universityIncludes Chinese University of Hong Kong. The Massachusetts Institute of Technology trained or employed 44 of the 187 AI major-prize winners, more than any other university. We compare doctorate and 5+ year employment links, career timing and current workplaces for 187 major-prize and 1,111 best-paper winners.
  2. Universities keep AI researchers for a median of 6 years, compared with 2 years at frontier labsIncludes Chinese University of Hong Kong. Across 15,811 dated work stays (6,529 people), median role tenure runs from 6 years at universities to 2 years at frontier labs. We compare all five organization types, starting decades from the 1970s to the 2020s, and the organizations that keep people longest.
  3. More people shaping AI studied at Massachusetts Institute of Technology than at any other institutionIncludes Massachusetts Institute of Technology. We estimate where the people shaping AI studied and where their careers led. Massachusetts Institute of Technology leads with 467 former students, including 291 recorded doctorates.
  4. The United States employs an estimated 48% of people shaping AIIncludes Chinese University of Hong Kong. We estimate where the people shaping AI work today, how universities, companies and research institutes split across countries, and how country shares have changed since the 1960s.

Colleagues

People with dated positions at the same organization in the same years.

Sources and credits

Photo: We have no published photo of Farzan Farnia yet. Send us a free-licensed photo.

  1. Farzan Farnia – CUHK CSE cse.cuhk.edu.hk. Checked 21 October 2026.
  2. A Minimax Approach to Supervised Learning papers.neurips.cc. Checked 21 October 2026.
  3. A Convex Duality Framework for GANs papers.neurips.cc. Checked 21 October 2026.
  4. Do GANs always have Nash equilibria? proceedings.mlr.press. Checked 21 October 2026.
  5. PhD Oral Defense: Farzan Farnia | Tse Lab at Stanford University tselab.stanford.edu. Checked 21 October 2026.
  6. Conditional Vendi Score: Prompt-Aware Diversity Evaluation for Generative AI Models and LLMs proceedings.mlr.press. Checked 21 October 2026.
  7. Farzan FARNIA - The Chinese University of Hong Kong research.cuhk.edu.hk. Checked 21 October 2026.