Research map / Optimization / Implicit bias and generalization
Position: Lottery Tickets Do Not Explain Overparameterization. How About Escape Dimensions?
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Position paper arguing that the lottery ticket analogy is misleading for explaining overparameterization and proposing escape dimensions as an alternative explanation.
Problem
Overparameterized neural networks are often explained via the lottery ticket hypothesis, which suggests larger networks succeed by containing well-initialized subnetworks; the paper argues this explanation is misleading.
Approach
Argues that subnetworks cannot be treated in isolation as the lottery analogy implies, and proposes escape dimensions as a better framework for understanding why overparameterization helps.
Experimental setup
Builds on: lottery ticket hypothesis
Closest papers by meaning
- Optimal Lottery Tickets via Subset Sum: Logarithmic Over-Parameterization is Sufficient (NeurIPS 2020 · similarity 0.90)
- Logarithmic Pruning is All You Need (NeurIPS 2020 · similarity 0.90)
- The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks (ICLR 2019 · similarity 0.89)
- Find A Winning Sign: Sign Is All We Need to Win the Lottery (ICLR 2025 · similarity 0.88)
- Weisfeiler and Leman Go Gambling: Why Expressive Lottery Tickets Win (ICML 2025 · similarity 0.88)
- Tilting the Odds at the Lottery: the Interplay of Overparameterisation and Curricula in Neural Networks (ICML 2024 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.