Sébastien Bubeck (born April 16, 1985) is a French-American

computer scientist A computer scientist is a scientist who specializes in the academic study of computer science. Computer scientists typically work on the theoretical side of computation. Although computer scientists can also focus their work and research on ...

and

mathematician A mathematician is someone who uses an extensive knowledge of mathematics in their work, typically to solve mathematical problems. Mathematicians are concerned with numbers, data, quantity, mathematical structure, structure, space, Mathematica ...

. He was

Microsoft Microsoft Corporation is an American multinational corporation and technology company, technology conglomerate headquartered in Redmond, Washington. Founded in 1975, the company became influential in the History of personal computers#The ear ...

's Vice President of Applied Research and led the Machine Learning Foundations group at

Microsoft Research Redmond Microsoft Research (MSR) is the research subsidiary of Microsoft. It was created in 1991 by Richard Rashid, Bill Gates and Nathan Myhrvold with the intent to advance state-of-the-art computing and solve difficult world problems through technologi ...

. Bubeck was formerly professor at

Princeton University Princeton University is a private university, private Ivy League research university in Princeton, New Jersey, United States. Founded in 1746 in Elizabeth, New Jersey, Elizabeth as the College of New Jersey, Princeton is the List of Colonial ...

and a researcher at the

University of California, Berkeley The University of California, Berkeley (UC Berkeley, Berkeley, Cal, or California), is a Public university, public Land-grant university, land-grant research university in Berkeley, California, United States. Founded in 1868 and named after t ...

. He is known for his contributions to online learning,

optimization Mathematical optimization (alternatively spelled ''optimisation'') or mathematical programming is the selection of a best element, with regard to some criteria, from some set of available alternatives. It is generally divided into two subfiel ...

and more recently studying

deep neural networks Deep learning is a subset of machine learning that focuses on utilizing multilayered neural networks to perform tasks such as classification, regression, and representation learning. The field takes inspiration from biological neuroscience a ...

, and in particular transformer models. Since 2024, he works for

OpenAI OpenAI, Inc. is an American artificial intelligence (AI) organization founded in December 2015 and headquartered in San Francisco, California. It aims to develop "safe and beneficial" artificial general intelligence (AGI), which it defines ...

Work

Bubeck's work spans a wide variety of topics in

machine learning Machine learning (ML) is a field of study in artificial intelligence concerned with the development and study of Computational statistics, statistical algorithms that can learn from data and generalise to unseen data, and thus perform Task ( ...

theoretical computer science Theoretical computer science is a subfield of computer science and mathematics that focuses on the Abstraction, abstract and mathematical foundations of computation. It is difficult to circumscribe the theoretical areas precisely. The Associati ...

and

artificial intelligence Artificial intelligence (AI) is the capability of computer, computational systems to perform tasks typically associated with human intelligence, such as learning, reasoning, problem-solving, perception, and decision-making. It is a field of re ...

. Some of his most notable contributions include developing minimax rate for multi-armed bandits, linear bandits, developing an optimal algorithm for bandit convex optimization, and solving long-standing problems in k-server and

metrical task systems Task systems are mathematical objects used to model the set of possible configurations of online algorithms. They were introduced by Borodin, Linial and Saks (1992) to model a variety of online problems. A task system determines a set of states a ...

. In regards to the mathematical theory of neural networks, Bubeck has both introduced and proved the ''law of robustness'' which links the number of parameters of a

neural network A neural network is a group of interconnected units called neurons that send signals to one another. Neurons can be either biological cells or signal pathways. While individual neurons are simple, many of them together in a network can perfor ...

and its regularity properties. Bubeck has also made contributions to

convex optimization Convex optimization is a subfield of mathematical optimization that studies the problem of minimizing convex functions over convex sets (or, equivalently, maximizing concave functions over convex sets). Many classes of convex optimization problems ...

, network analysis, and

information theory Information theory is the mathematical study of the quantification (science), quantification, Data storage, storage, and telecommunications, communication of information. The field was established and formalized by Claude Shannon in the 1940s, ...

. Bubeck's papers have over 25,000 citations to date. Prior to joining Microsoft Research, Bubeck was an assistant professor at

in the Department of Operations Research and Financial Engineering. He received his PhD from the Lille 1 University of Science and Technology, and also studied at the Ecole Normale Supérieure de Cachan. Bubeck is the author of the book ''Convex optimization: Algorithms and complexity'' (2015). He has also been on the editorial board of several scientific journals and conferences, including the ''Journal of the ACM'' and ''

Neural Information Processing Systems The Conference and Workshop on Neural Information Processing Systems (abbreviated as NeurIPS and formerly NIPS) is a machine learning and computational neuroscience conference held every December. Along with ICLR and ICML, it is one of the three ...

'' (NeurIPS) and was program committee chair for the 2018 Conference on Learning Theory (COLT) In 2023, Bubeck and his collaborators published a paper that claimed to observe "sparks of

artificial general intelligence Artificial general intelligence (AGI)—sometimes called human‑level intelligence AI—is a type of artificial intelligence that would match or surpass human capabilities across virtually all cognitive tasks. Some researchers argue that sta ...

" in an early version of

GPT-4 Generative Pre-trained Transformer 4 (GPT-4) is a multimodal large language model trained and created by OpenAI and the fourth in its series of GPT foundation models. It was launched on March 14, 2023, and made publicly available via the p ...

, a large language model developed by

. The paper presented examples of GPT-4 performing tasks across various domains and modalities, such as mathematics, coding, vision, medicine, and law. The paper sparked wide interest and debate in the scientific community and the popular media, as it challenged the conventional understanding of learning and cognition in AI systems. Bubeck also investigated the potential use of GPT-4 as an AI chatbot for medicine in a paper that evaluated the strengths, weaknesses, and ethical issues of relying on such a tool for medical purposes In October 2024, Bubeck left Microsoft to join

Honors and awards

Bubeck has received numerous honors and awards for his work, including the

Alfred P. Sloan Research Fellowship The Sloan Research Fellowships are awarded annually by the Alfred P. Sloan Foundation since 1955 to "provide support and recognition to early-career scientists and scholars". This program is one of the oldest of its kind in the United States. ...

in Computer Science in 2015, and Best Paper Awards at the Conference on Learning Theory (COLT) in 2016,

(NeurIPS) in 2018 and 2021 and in the ACM Symposium on Theory of Computing (STOC) 2023. He has also received the Jacques Neveu prize for the best French PhD in Probability/Statistics, the runner-up prize in AfIA's 2011 French AI thesis awards, and one of the two second prizes in the 2010 Gilles Kahn prize for a French PhD in computer science.

Selected publications

* Minimax policies for adversarial and stochastic bandits (2009), with Jean-Yves Audibert. * Best arm identification in multi-armed bandits (2010), with Jean-Yves Audibert and Rémi Munos. * Kernel-based methods for bandit convex optimization (2017), with Yin Tat Lee and Ronen Eldan. * A universal law of robustness via isoperimetry (2020), with Mark Sellke. * K-server via multiscale entropic regularization (2018), with Michael B. Cohen, Yin Tat Lee, James R. Lee, and Aleksander Madry. * Competitively chasing convex bodies (2019), with Yin Tat Lee, Yuanzhi Li, and Mark Sellke. * Regret analysis of stochastic and nonstochastic multi-armed bandit problems (2012), with

Nicolò Cesa-Bianchi Nicolò Cesa-Bianchi () is an Italian computer scientist and Professor of Computer Science at the Department of Computer Science of the University of Milan. He is a researcher in the field of machine learning Machine learning (ML) is a ...

References

{{DEFAULTSORT:Bubeck, Sébastien 1985 births Living people French computer scientists French mathematicians American computer scientists American mathematicians Lille University of Science and Technology alumni Princeton University faculty Microsoft Research people OpenAI people