PeopleofSuperintelligence

Researcher

Nikolas Nüsken

Nikolas Nüsken is a mathematician at King's College London whose work on stochastic analysis and optimal control supports machine-learning methods for sampling and diffusion models. After a PhD at Imperial College London in 2018 under Greg Pavliotis and a postdoctoral post at the University of Potsdam, he joined King's in 2022. A 2021 paper with Lorenz Richter and Leon Sallandt on high-dimensional parabolic PDEs in the tensor train format received an Outstanding Paper Honorable Mention at ICML. His other work covers Stein variational gradient descent with Michiel Renger in 2021, Controlled Monte Carlo Diffusions with Francisco Vargas, Shreyas Padhy and Denis Blessing in 2023, and conditioning diffusions with Malliavin calculus at ICML 2025. He is a Lecturer in Mathematical Data Science in the Department of Mathematics at King's College London.

Timeline

  1. 2018

    A doctorate at Imperial

    He obtains his PhD from Imperial College London in 2018 under the supervision of Professor Greg Pavliotis.

  2. 2020

    Neural solvers for HJB equations

    In May 2020, he and Lorenz Richter post a paper on solving high-dimensional Hamilton-Jacobi-Bellman PDEs with neural networks, drawing on the theory of controlled diffusions and measures on path space. They build a framework on divergences between path measures and propose the log-variance divergence, with applications in importance sampling and rare event simulation.

  3. 2021

    An ICML honorable mention

    With Lorenz Richter and Leon Sallandt, he writes "Solving high-dimensional parabolic PDEs using the tensor train format", which pairs backward stochastic differential equations with regression in the tensor format. The paper receives an Outstanding Paper Honorable Mention at ICML in 2021.

  4. 2021–2022

    Stein variational gradient descent

    In February 2021, he and Michiel Renger post a paper on Stein variational gradient descent, a class of Bayesian inference methods built on interacting particle systems. They link variational inference and Markov chain Monte Carlo through gradient flow structures and large-deviation principles, and Foundations of Data Science publishes the paper online in December 2022.

  5. 2022

    Joining King's College London

    After a short stay at the Alan Turing Institute, he works as a postdoctoral researcher at the University of Potsdam within the Collaborative Research Centre "Scaling Cascades in Complex Systems". In 2022, he joins King's College London.

  6. 2023

    Controlled Monte Carlo Diffusions

    In July 2023, he and Francisco Vargas, Shreyas Padhy and Denis Blessing post "Transport meets Variational Inference: Controlled Monte Carlo Diffusions", which connects optimal transport and variational inference and introduces a score-based annealing sampler for Bayesian computation. He presents the work at an Imperial College London seminar in October 2023.

  7. 2025

    Conditioning diffusions with Malliavin calculus

    At ICML 2025, he coauthors "Conditioning Diffusions Using Malliavin Calculus" with Jakiw Pidstrigach, Elizabeth Louise Baker, Carles Domingo-Enrich and George Deligiannidis. The paper introduces a framework built on Malliavin calculus and a generalisation of the Tweedie score formula, which handles singular terminal rewards such as those in diffusion bridges.

  8. As of October 2026

    Lecturer in mathematical data science

    He is a Lecturer in Mathematical Data Science in the Department of Mathematics at King's College London, with research interests in computational Bayesian inference, optimal control and optimal transport, and kernel methods. In December 2025, he posts "Control Consistency Losses for Diffusion Bridges" with Samuel Howard and Jakiw Pidstrigach, and it is a poster at ICML 2026 in July 2026.

Career

  1. NowKing's College LondonLecturer in Mathematical Data Science
  2. Until 2022University of PotsdamPostdoctoral Researcher

Education

  1. 2018Imperial College LondonPhD

From the biography.

  1. The median h-index is 48 at universities and 14 at frontier AI labsIncludes King's College London. The median h-index is 48 at universities and 14 at frontier AI labs. We compare publishing careers, recent citations, dual appointments and job moves across 2,420 university researchers and 308 frontier-lab researchers.
  2. Companies supply the most prior-work hires at 7 of the 9 frontier AI labsIncludes King's College London. Our estimates of the organizations where 550 people worked directly before joining 9 frontier AI labs across 620 career transitions, comparing the main feeder organizations per lab.
  3. More people shaping AI studied at Massachusetts Institute of Technology than at any other institutionIncludes King's College London. 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 company share of AI researchers’ top papers was 31% in 2020–2024 and 8% in 1970–1974Includes King's College London. 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 Nikolas Nüsken yet. Send us a free-licensed photo.

  1. Nikolas Nüsken - King's College London kcl.ac.uk. Checked 22 October 2026.
  2. Solving high-dimensional Hamilton-Jacobi-Bellman PDEs using neural networks: perspectives from the theory of controlled diffusions and measures on path space arxiv.org. Checked 22 October 2026.
  3. ICML 2021 Awards icml.cc. Checked 22 October 2026.
  4. Solving high-dimensional parabolic PDEs using the tensor train format proceedings.mlr.press. Checked 22 October 2026.
  5. Stein Variational Gradient Descent: many-particle and long-time asymptotics arxiv.org. Checked 22 October 2026.
  6. Stein variational gradient descent: Many-particle and long-time asymptotics aimsciences.org. Checked 22 October 2026.
  7. Transport meets Variational Inference: Controlled Monte Carlo Diffusions arxiv.org. Checked 22 October 2026.
  8. Nikolas Nusken (King's College London) and Lachlan Astfalck (Univ. Western Australia) imperial.ac.uk. Checked 22 October 2026.
  9. Conditioning Diffusions Using Malliavin Calculus proceedings.mlr.press. Checked 22 October 2026.
  10. Control Consistency Losses for Diffusion Bridges (ICML 2026) icml.cc. Checked 22 October 2026.
  11. Control Consistency Losses for Diffusion Bridges arxiv.org. Checked 22 October 2026.