Strange Geometric Shapes Found Inside AIs — Tom McGrath - YouTube
Tom McGrath is co-founder and Chief Scientist at Goodfire, and a former Google DeepMind researcher. He joins Tim Scarfe to ask what neural networks actually ...
Tom McGrath is co-founder and Chief Scientist at Goodfire, and a former Google DeepMind researcher. He joins Tim Scarfe to ask what neural networks actually learn, whether their internal representations converge on structures in the world, and whether interpretability can extract new scientific knowledge rather than merely explain model outputs. Beginning with AlphaZero and learned modularity, the conversation moves into neural geometry: concept manifolds, reusable computation inside Llama, and why activation steering can fail when it pushes a model off-manifold. McGrath then makes the case…
saved by
related reading
- AI Is Learning at the Wrong Level of Abstraction — Matthieu Wyartyoutube.com
- Recursive Language Models — Alex Zhang, MIT PhDyoutube.com
- Why Frontier AI Labs Fight to Hide Chain of Thought — Ilia Shumailov & Alexander Panfilovyoutube.com
- AI researchers debate how close we are to recursive self-improvementyoutube.com
- Why Sam Altman is against a single superintelligenceyoutube.com
- AI researchers talk continual learning over a game of HORSEyoutube.com
- RLMs are Compositional Generalizers with Alex Zhang from MITyoutube.com
- How AI is Changing Programmingyoutube.com
- We have 3 years to solve alignment before superintelligenceyoutube.com
- Why Inference Matters 12x More Than Training | YC Paper Clubyoutube.com
- What if not Circuits? — LessWronglesswrong.com
- The World Inside Neural Networksgoodfire.ai