PeopleofSuperintelligence

Researcher

Kartik Chandra

Kartik Chandra is a computer scientist whose research asks how minds understand other minds, drawing on programming languages theory, computer graphics and AI. A Stanford graduate in computer science, he became a Hertz Fellow and received a Paul & Daisy Soros Fellowship in 2021 to pursue a PhD at MIT. With Audrey Xie, Jonathan Ragan-Kelley and Erik Meijer, he won a NeurIPS 2022 Outstanding Paper award for "Gradient Descent: The Ultimate Optimizer", a method that lets gradient descent tune its own hyperparameters. In 2025, a paper he wrote with Tony Chen, Josh Tenenbaum and Ragan-Kelley introduced memo, a programming language for modeling theory of mind, and received the ACM SIGPLAN Distinguished Paper Award; in 2026, he and three coauthors showed in a Bayesian model that sycophantic chatbots can drive even an ideal Bayes-rational user into delusional spiraling. He is a final-year PhD student at MIT CSAIL, advised by Jonathan Ragan-Kelley and Josh Tenenbaum.

Timeline

  1. By 2021

    Stanford computer science degree

    He earns a B.S. in computer science with honors from Stanford, graduating in 2021. His undergraduate research with professors Ras Bodik, Alex Aiken, Erik Meijer and Gregory Valiant leads to publications and open-source software, and he is a 2020 Barry Goldwater Scholar.

  2. 2021

    Hertz and Soros fellowships

    In 2021, he becomes a Hertz Fellow and receives a Paul & Daisy Soros Fellowship for New Americans to pursue a PhD in electrical engineering and computer science at MIT.

  3. 2022

    An outstanding paper at NeurIPS

    He, Audrey Xie, Jonathan Ragan-Kelley and Erik Meijer win a NeurIPS 2022 Outstanding Paper award for "Gradient Descent: The Ultimate Optimizer". The paper modifies backpropagation so that gradient descent computes hypergradients and tunes its own hyperparameters, such as step size, and the method can be stacked recursively.

  4. 2022–2024

    Leading MIT's application assistance program

    He leads MIT's Graduate Application Assistance Program from 2022 to 2024.

  5. 2023

    Storytelling as inverse inverse planning

    In 2023, he publishes "Acting as Inverse Inverse Planning" at SIGGRAPH with Tzu-Mao Li, Josh Tenenbaum and Jonathan Ragan-Kelley. The paper optimizes animations against the experiences of simulated audience members, treating storytelling as choosing actions to manipulate an inverse planner's inferences.

  6. 2025

    The memo programming language

    In 2025, he, Tony Chen, Josh Tenenbaum and Jonathan Ragan-Kelley publish memo, a probabilistic programming language with specialized syntax and semantics for theory of mind, and the paper receives the ACM SIGPLAN Distinguished Paper Award. Its models can take up to three times fewer lines of code and, in some cases, run 3,000 times faster, or more.

  7. February 2026

    Sycophancy and delusional spiraling

    In February 2026, he, Max Kleiman-Weiner, Jonathan Ragan-Kelley and Josh Tenenbaum post "Sycophantic Chatbots Cause Delusional Spiraling, Even in Ideal Bayesians". Their Bayesian model of a user talking to a chatbot shows that sycophancy plays a causal role, and the effect persists when chatbots are prevented from hallucinating and users are told that a model may be sycophantic.

  8. As of October 2026

    PhD student at MIT CSAIL

    He is a final-year PhD student at MIT CSAIL, advised by Jonathan Ragan-Kelley and Josh Tenenbaum, and studies the computational basis of theory of mind. In September 2026, he posts "Theories of Mind as Domain-Specific Languages of Thought" with Ragan-Kelley, Tenenbaum and Rebecca Saxe, which proposes formalizing theories of mind as domain-specialized programming languages.

Career

  1. 2022–2024Massachusetts Institute of TechnologyProgram Lead

Education

  1. 2021Stanford UniversityB.S. in Computer Science
  2. From 2021Massachusetts Institute of TechnologyPhD

From the biography.

  1. 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.
  2. The Massachusetts Institute of Technology trained or employed 44 of the 187 AI major-prize winners, more than any other universityIncludes Massachusetts Institute of Technology. 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.
  3. Google had the most people shaping AI every year from 2014 to 2025Includes Massachusetts Institute of Technology. Our estimates of which organizations employed the most people shaping AI each year from 1955 to 2025, tracing the shift from university labs to companies.
  4. The company share of AI researchers’ top papers was 31% in 2020–2024 and 8% in 1970–1974Includes Massachusetts Institute of Technology. We estimate where the people shaping AI worked when their most-cited papers appeared: the company share was 31% in 2020–2024 and 8% in 1970–1974, and Google led all organizations in 2020–2024.

Colleagues

People named in this biography, and people with dated positions at the same organization in the same years.

Sources and credits

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

  1. Kartik Chandra - Hertz Foundation hertzfoundation.org. Checked 21 October 2026.
  2. Kartik Chandra – Paul & Daisy Soros Fellowships for New Americans pdsoros.org. Checked 21 October 2026.
  3. Three from MIT receive 2021 Hertz Foundation Fellowships | MIT News news.mit.edu. Checked 21 October 2026.
  4. NeurIPS 2022 Awards neurips.cc. Checked 21 October 2026.
  5. #NeurIPS2022 outstanding paper – Gradient descent: the ultimate optimizer aihub.org. Checked 21 October 2026.
  6. NeurIPS 2022 Announces Its Outstanding Main Track Papers, Outstanding Dataset & Benchmark Papers, and Test of Time Award syncedreview.com. Checked 21 October 2026.
  7. Kartik Chandra cs.stanford.edu. Checked 21 October 2026.
  8. Acting as Inverse Inverse Planning arxiv.org. Checked 21 October 2026.
  9. memo Helps Researchers Model How Minds Think About Other Minds sqi.mit.edu. Checked 21 October 2026.
  10. A Domain-Specific Probabilistic Programming Language for Reasoning about Reasoning (or: a memo on memo) 2025.splashcon.org. Checked 21 October 2026.
  11. Sycophantic Chatbots Cause Delusional Spiraling, Even in Ideal Bayesians arxiv.org. Checked 21 October 2026.
  12. Kartik Chandra – Computational Cognitive Science Lab cocosci.mit.edu. Checked 21 October 2026.
  13. Theories of Mind as Domain-Specific Languages of Thought arxiv.org. Checked 21 October 2026.