References & Citations
Computer Science > Machine Learning
Title: Insights into the Lottery Ticket Hypothesis and Iterative Magnitude Pruning
(Submitted on 22 Mar 2024 (v1), last revised 27 Mar 2024 (this version, v2))
Abstract: Lottery ticket hypothesis for deep neural networks emphasizes the importance of initialization used to re-train the sparser networks obtained using the iterative magnitude pruning process. An explanation for why the specific initialization proposed by the lottery ticket hypothesis tends to work better in terms of generalization (and training) performance has been lacking. Moreover, the underlying principles in iterative magnitude pruning, like the pruning of smaller magnitude weights and the role of the iterative process, lack full understanding and explanation. In this work, we attempt to provide insights into these phenomena by empirically studying the volume/geometry and loss landscape characteristics of the solutions obtained at various stages of the iterative magnitude pruning process.
Submission history
From: Tausifa Jan Saleem [view email][v1] Fri, 22 Mar 2024 08:11:14 GMT (25273kb,D)
[v2] Wed, 27 Mar 2024 10:47:24 GMT (25273kb,D)
Link back to: arXiv, form interface, contact.