Profile avatar
onnoeberhard.com
PhD Student in Tübingen (MPI-IS & Uni Tü), interested in reinforcement learning. Ex research intern at Google Research. https://onnoeberhard.com/
11 posts 426 followers 208 following
Conversation Starter
comment in response to post
Actually, this argument applies generally to functions with connected sublevel sets, so it is not specific to star-shaped sets. (also I meant "f(z) < f(y)" above)
comment in response to post
It seems like a local minimum should imply a global minimum with such functions. Otherwise, if z is a local minimum (f(z) > f(y) in neighborhood around z), then the sublevel set {x: f(x) < f(z)+eps} includes both z and the global minimum but not the neighborhood of z. This set is not star-shaped.
comment in response to post
I recently learned that this book was not even written by Feynman (even thought it says so on the cover)! If you have 3 hours.. youtu.be/TwKpj2ISQAc
comment in response to post
There is, the extension is called “Latex Workshop”
comment in response to post
Sounds interesting, looking forward to it!
comment in response to post
What is this “forthcoming book”? 👀
comment in response to post
Could this be related to the options framework?
comment in response to post
Could you please add me? :)
comment in response to post
Would be happy to be added :)
comment in response to post
It also implies the existence of Amnesia (Arthur-Merlin). I don’t remember why I said that.
comment in response to post
Absolutely! Are those also bicycle trajectories?