14678db82874f1456031fcc05a3afaf6-AuthorFeedback.pdf
–Neural Information Processing Systems
Regarding the comment in lines 198-202, we apologize for any confusion. Regarding variance in experiments, we haveobserved high variance is not enough for the algorithm to get "lucky".31 In fact, we observe that thebest performance of Greedy is often worse than theworst performance of IHT.
Neural Information Processing Systems
Feb-11-2026, 13:06:05 GMT