Our news
-
Humanity meets AI
feb 27-28 How might AI influence our understanding of humanity, morality, and meaning-making? Prof. Kreiman participated in a discussion panel with Prof. Dashi-Velez and Dr. Goodwin organized by the Harvard Divinity School The discussion focused on fascinating issues at the intersection of technology development, well-being, morality, and how to think about the role of humanity…
-
AAAI 2025: The indoor training effect
Feb 27, 2025 Presentation at AAAI 2025 Serena Bono, Spandan Madan and colleagues present their work titled “The indoor training effect: unexpected gains from distribution shifts in the transition function”. This work reports surprising findings where algorithms trained under noisy conditions can generalize better than algorithms trained under noise-free conditions.
-
The indoor training effect
feb 17, 2025 Scientists thought this would make AI worse but it made it smarter Note by Adam Zewe, MIT, about the work by Serena Bono and Spandan Madan in AAAI 2025 (see PDF here). SciTechDaily See more related work from the lab:
-
Brains Minds and Machines Summer Course 2025
August 3, 2025 — august 24, 2025 Apply now to join the BMM Summer School The goal of this course is to help produce a community of leaders that is equally knowledgeable in neuroscience, cognitive science, and computer science and will lead the scientific understanding of intelligence and the development of true biologically inspired AI.
-
Using computational models to improve visual learning
jan 30, 2025 Morgan Talbot presenting at Vision Journal Club Morgan will be presenting the paper “L-WISE: Boosting Human Image Category Learning Through Model-Based Image Selection and Enhancement.” For a concise summary, please see the project website. The paper explores ways to enhance visual category learning in humans by applying adversarially trained ANNs as models of…
-
How do neurons encode the order of events?
jan 29, 2025 Prefrontal and medial temporal neurons encode ordinal information of event sequences in humans Elisa Pavarino will be presenting her work jointly with Jie Zheng and Ueli Rurishauser.
-
Generalization in ML: the indoor training effect
Jan 29, 2025 Is it better to train under noisy conditions to learn to generalize? New work by Serena Bono and Spandan Madan shed light on the mechanisms underlying generalization in reinforcement learning agents. This work led to discovering the indoor training effect whereby agents can improve their generalization performance under certain circumstances when trained…
-
Generalization in models of vision
Jan 8, 2025 Evaluating how brains generalize: data from macaque monkeys reveals flaws in deep neural networks Anne Manning, Harvard John A. Paulson School of Engineering and Applied Sciences, reports on the work of Spandan Madan, Will Xiao and colleagues presented recently in NeurIPS. This work sheds light on the key Achilles heel of current…
-
Generalization is the Achilles heel of AI
dec 18, 2024 Kempner All Hands Meeting Prof. Kreiman gave an All Hands meeting at the Kempner Institute for the Study of Artificial and Natural Intelligence at Harvard. The talk focused on the challenges of building AI algorithms that can generalize to novel data and situations. Read more about some of the research presented in…
-
Kreiman lab presentations at NeurIPS 2024
dec 10-15, 2024 Neural Information Processing Systems 2024 Several members of the Kreiman lab presented their cutting-edge research at NeurIPS 2024.