Research map / Architectures and efficiency / Neural architecture search
Per-Architecture Training-Free Metric Optimization for Neural Architecture Search
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.
Proposes a per-architecture method that combines multiple training-free metrics to estimate network performance for NAS without training.
Problem
Training-free metrics for NAS each capture only partial architectural features and vary in estimation ability across tasks, so combining them is needed to reduce NAS deployment cost.
Approach
Optimizes the combination of multiple training-free metrics separately for each architecture rather than using one global combination across the search space (inferred from the title; the abstract is truncated).
Closest papers by meaning
- Unifying and Boosting Gradient-Based Training-Free Neural Architecture Search (NeurIPS 2022 · similarity 0.94)
- Robustifying and Boosting Training-Free Neural Architecture Search (ICLR 2024 · similarity 0.92)
- Neural Architecture Search on ImageNet in Four GPU Hours: A Theoretically Inspired Perspective (ICLR 2021 · similarity 0.92)
- einspace: Searching for Neural Architectures from Fundamental Operations (NeurIPS 2024 · similarity 0.90)
- SWAP-NAS: Sample-Wise Activation Patterns for Ultra-fast NAS (ICLR 2024 · similarity 0.90)
- Insights from the Use of Previously Unseen Neural Architecture Search Datasets (CVPR 2024 · similarity 0.90)
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.