Closed negrinho closed 5 years ago
Thanks @negrinho you are right I counted twice in cases like these, I am going to correct them ASAP.
Practical and Consistent Estimation of f-Divergences Paul Rubenstein (MPI for IS) · Olivier Bousquet (Google Brain (Zurich)) · Josip Djolonga (Google Research, Brain Team) · Carlos Riquelme (Google Brain) · Ilya Tolstikhin (MPI for Intelligent Systems)
Unless I'm missing something, I believe that the stats you published are not correct. Please see script and results. I got 83 for both MIT and Stanford, 80 for CMU, and 56 for Berkeley.