blog
thoughts, tutorials, and musings
-
When LLMs Play Games: The Cooperation Secret
What happens when language models face the Prisoner's Dilemma?
-
Your AI Knows More Than It Can Do
Understanding the systematic gap between what language models can describe and what they can actually do
-
The Car Wash Problem: When LLMs miss the obvious
Exploring how LLMs fail at pragmatic reasoning and why this matters for AI safety
-
Isn't Deactivating Neurons So Good?
Conducting ablation experiments in neural networks