videos.
First Principles: Razvan Pascanu
LLMs vs. LTMs
First Principles: Sander Dieleman
Tree-vial Pursuits: How Humble Decision Trees Still Outsmart the Generative Giants
Beyond Bigger Models: From StarCraft to Representation
COMARL Virtual Seminar
Meta-Learning and Neural Processes
Designing Learning Dynamics
Sequences and Recurrent Networks
Generative Query Networks and Neural Processes
Symbolic Representation Learning
press.
2026
- alphaXiv Why Large Language Models Fail at Tabular Prediction Summary of my paper with Wojciech Czarnecki Aug ↗
- IEEE Spectrum Large Tabular Models Excel Where LLMs Fail Feature on Fundamental and NEXUS Jul ↗
- TechCrunch Fundamental raises $255 million Series A with a new take on big data analysis Fundamental's launch and Series A Feb ↗
- Financial Times AI venture Fundamental secures $1.2bn valuation and Amazon deal Fundamental's valuation and Amazon partnership Feb ↗
- Salesforce Ventures Bringing AI to Structured Data Why they invested in Fundamental Feb ↗
- neolabs.fyi Fundamental AI lab directory ↗
2021
2018
- Inverse In 2018, Google DeepMind imbued A.I. with human-like sight and imagination Generative Query Networks among the top AI of 2018 Dec ↗
- The Next Web DeepMind's AI learned to turn flat images into 3D scenes Generative Query Networks, published in Science Jun ↗
- MIT Technology Review A computer program that learns to “imagine” the world Generative Query Networks, published in Science Jun ↗
- secrss.com GQN extensions and Neural Processes Chinese-language coverage of my Neural Processes papers ↗
- Kaspar Märtens Kaspar Märtens Neural Processes as distributions over functions Explainer of my Neural Processes papers ↗
2016
- The Morning Paper The Morning Paper Towards deep symbolic reinforcement learning Review of my first-author PhD paper Oct ↗
- The Verge These are three of the biggest problems facing today's AI Features our deep symbolic reinforcement learning work Oct ↗
-
New Scientist Deep symbolic reinforcement learning Two-page feature on our deep symbolic reinforcement learning work Oct
-
The Guardian Deep symbolic reinforcement learning Coverage of our deep symbolic reinforcement learning work Sep