paper-with-me

홈 › Papers

Not All Lotteries Are Made Equal

2022-06-16 · Surya Kant Sahu, Sai Mitheran, Somya Suhans Mahapatra

The Lottery Ticket Hypothesis (LTH) states that for a reasonably sized neural network, a sub-network within the same network yields no less performance than the dense counterpart when trained from the same initialization. This work investigates the relation between model size and the ease of finding these sparse sub-networks. We show through experiments that, surprisingly, under a finite budget, smaller models benefit more from Ticket Search (TS).

📄 PDF Abstract BibTeX arXiv:2206.08175

Code (0)

등록된 구현이 없습니다.

Tasks

AllRelationTicket Search

Similar Papers 제목 키워드 기반

Making school choice lotteries transparent

2025-01-08 · Lingbo Huang, Jun Zhang

Lotteries are commonly employed in school choice to fairly resolve priority ties; however, current practices leave students uninformed about their lottery outcomes when submitting preferences. This paper advocates for re…

Decision Making

Regret theory, Allais' Paradox, and Savage's omelet

2023-01-06 · Vardan G. Bardakhchyan, Armen E. Allahverdyan

We study a sufficiently general regret criterion for choosing between two probabilistic lotteries. For independent lotteries, the criterion is consistent with stochastic dominance and can be made transitive by a unique c…

Not All Lotteries Are Made Equal

2022-01-17 · ICLR Track Blog 2022 5 · Anonymous

The Lottery Ticket Hypothesis (LTH) states that for a reasonably sized neural network, there exists a subnetwork within the same network that yields no less performance than the dense counterpart when trained from the sa…

All

The Near Miss Effect and the Framing of Lotteries

2021-07-06 · Michael Crystal

We present a framework for analyzing the near miss effect in lotteries. A decision maker (DM) facing a lottery, falsely interprets losing outcomes that are close to winning ones, as a sign that success is within reach. A…

Robust AI Evaluation through Maximal Lotteries

2026-02-24 · Hadi Khalaf, Serena L. Wang, Daniel Halpern, Itai Shapira 외 arxiv

The standard way to evaluate language models on subjective tasks is through pairwise comparisons: an annotator chooses the "better" of two responses to a prompt. Leaderboards aggregate these comparisons into a single Bra…