PeopleofSuperintelligence

Researcher

Alex Lamb

Alex Lamb is a machine-learning researcher whose work covers demand forecasting, Japanese historical documents and AI representations. He studied computer science and applied mathematics and statistics at Johns Hopkins from 2011 to 2013. At Amazon, he developed demand-forecasting algorithms and systems, then earned an M.Sc. in computer science from the University of Montreal in 2018. His coauthored work includes the Kuzushiji datasets, KuroNet, Manifold Mixup and AC-State; he completed a 2020 PhD at Université de Montréal under Yoshua Bengio and received the 2020 Twitch Research Fellowship. Since June 2025, he is an assistant professor and PI of Lamb Lab at Tsinghua University’s College of AI, after working at Microsoft Research NYC.

Works
84
Citations
6,530
h-index
22
Badges

Publication numbers from OpenAlex, as of 27 October 2026.

Timeline

  1. 2011–2013

    Johns Hopkins and Twitter classifiers

    He earns a B.Sc. in computer science and applied mathematics and statistics at Johns Hopkins University from 2011 to 2013. As an undergraduate there, he develops text classifiers for Twitter to measure and monitor flu outbreaks.

  2. 2013–2015

    Demand forecasting at Amazon

    From 2013 to 2015, he works as an applied research scientist at Amazon and develops algorithms and systems for demand forecasting.

  3. 2016–2018

    An M.Sc. in Montreal

    In 2018, he completes an M.Sc. in computer science at the University of Montreal with Aaron Courville.

  4. 2018–2019

    Kuzushiji datasets and KuroNet

    In 2018, he coauthors Kuzushiji-MNIST, Kuzushiji-49 and Kuzushiji-Kanji for cursive Japanese. In 2019, he coauthors KuroNet, an end-to-end model that recognizes an entire page of pre-modern Japanese text.

  5. 2019

    Manifold Mixup

    In 2019, he coauthors Manifold Mixup, a regularizer that uses interpolated hidden representations to make neural-network decisions less confident. The paper improves strong baselines in supervised learning, robustness to single-step adversarial attacks and test log-likelihood.

  6. 2020

    A PhD at Université de Montréal

    In 2020, he completes a PhD in computer science at Université de Montréal, advised by Yoshua Bengio.

  7. 2020

    A Twitch research fellowship

    In 2020, Twitch names him among its Research Fellowship award winners. Each winner receives $10,000 for academic research and an offer to connect with a team at Twitch for a full-time paid internship at Twitch HQ.

  8. February 2023

    Agent-centric state discovery

    In February 2023, he coauthors AC-State, which discovers a minimal control-endogenous latent state from observations. It enables localization, exploration and navigation without reward or demonstrations.

  9. Since June 2025

    Assistant professor and lab PI

    In June 2025, he joins Tsinghua University’s College of AI as an assistant professor and PI of Lamb Lab. Before this move, he works as a deep-learning research scientist at Microsoft Research NYC and a senior researcher in Microsoft’s AI Frontiers group. Lamb Lab develops learning systems that rapidly acquire knowledge and adapt to new environments.

Career

  1. Since 2025Tsinghua UniversityAssistant professor, PI
  2. 2022–2025MicrosoftDeep-learning research scientist
  3. 2013–2015AmazonApplied Research Scientist

Education

  1. 2020Université de MontréalPhD
  2. 2018Université de MontréalM.Sc.
  3. 2013Johns Hopkins UniversityB.Sc.

From the biography and OpenAlex affiliations (the years of papers that list the organization), so some years are approximate.

Checked career sources: collegeai.tsinghua.edu.cn, collegeai.tsinghua.edu.cn.

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 Alex Lamb yet. Send us a free-licensed photo.

  1. Home sites.google.com. Checked 26 October 2026.
  2. [1812.01718] Deep Learning for Classical Japanese Literature arxiv.org. Checked 26 October 2026.
  3. [1910.09433] KuroNet: Pre-Modern Japanese Kuzushiji Character Recognition with Deep Learning arxiv.org. Checked 26 October 2026.
  4. Manifold Mixup: Better Representations by Interpolating Hidden States proceedings.mlr.press. Checked 26 October 2026.
  5. Alex LAMB-College of Al, Tsinghua University collegeai.tsinghua.edu.cn. Checked 26 October 2026.
  6. Introducing our 2020 Twitch Research Fellows blog.twitch.tv. Checked 26 October 2026.
  7. Guaranteed Discovery of Control-Endogenous Latent States with Multi-Step Inverse Models - Microsoft Research microsoft.com. Checked 26 October 2026.
  8. ECE Seminar - Discovering Agent-Centric Latent States in Theory and in Practice | University Event Calendar - The Hong Kong University of Science and Technology calendar.hkust.edu.hk. Checked 26 October 2026.
  9. Research Groups-College of Al, Tsinghua University collegeai.tsinghua.edu.cn. Checked 26 October 2026.