
23 - Mechanistic Anomaly Detection with Mark Xu
AXRP - the AI X-risk Research Podcast
How to Maximize the Entropy of Your Probability Distribution
In Bayesian reasoning, there's this idea of a probability distribution over some like set of interest and you don't know what your probability distribution is. And basically you say, well, given that I only know these things, I want to maximize the entropy of my probability distribution. So if we condition on C being 5050, the maximum entropy distribution over A and B has like A and B correlated a little bit but also it raises the probability of A relative to,. um, relative to 5050. Okay. Then you can just add those wires to each other where they're all going to be correlated. But then if you want the best possible correlation between two wires, which will push the
00:00
Transcript
Play full episode
Remember Everything You Learn from Podcasts
Save insights instantly, chat with episodes, and build lasting knowledge - all powered by AI.