159 citations · 159 across the 2 of their papers we have counts for
2 papers
cs.LG2024
Opening the Black Box: predicting the trainability of deep neural networks with reconstruction entropy
Yanick Thurn, Ro Jefferson, Johanna Erdmenger
An important challenge in machine learning is to predict the initial conditions under which a given neural network will be trainable. We present a method for predicting the trainab…
hep-th2007★ 159 cited
AdS/CFT with Flavour in Electric and Magnetic Kalb-Ramond Fields
Johanna Erdmenger, Rene Meyer, Jonathan P. Shock
We investigate gauge/gravity duals with flavour for which pure-gauge Kalb-Ramond B fields are turned on in the background, into which a D7 brane probe is embedded. First we conside…