
Episode 15: Martín Arjovsky, INRIA, on benchmarks for robustness and geometric information theory
Generally Intelligent
Is the Entropy of a Negative Infinite?
It just does not make sense to measure most of the things that people want to measure. Its just flat t wrong when you have geometry. There are much better tors ther these things are call metrican coveting numbers reova, they ar caye very poorly to o dimensional spaces. I think that the analogy cow the vasertin distance fixes a lot of problems that ca divergencies cav and geometry simple in the sense that the vacertain distance between two istivisions that are like lines. If you have lines that are far apart, the ascertine distance will be bigger than if they're very close,. while theca divergenci will be infinitely moins.
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.