arXiv cs.AI· Konstanty Subbotko·· 13 小时前AI 评分18
MIDAS：基于自注意力与分块输入的输入特定可微分架构搜索
MIDAS: Mosaic Input-Specific Differentiable Architecture Search
AI 导读
MIDAS 通过自注意力将 DARTS 的静态架构参数替换为输入特定的动态参数，并按激活图的空间分块分别计算架构选择，同时引入无参数、拓扑感知的搜索空间建模双边缘节点拓扑。
正文 · 原文
Abstract:Differentiable Neural Architecture Search (NAS) provides efficient, gradient-based methods for automatically designing neural networks, yet its adoption remains limited in practice. We present MIDAS, an approach that modernizes DARTS by replacing static architecture parameters with dynamic, input-specific parameters computed via self-attention. To improve robustness, MIDAS (i) localizes the architecture selection by computing it separately for each spatial patch of the activation map, and (ii) introduces a parameter-free, topology-aware search space that models the two-edge node topology during search. We evaluate MIDAS on the DARTS, NAS-Bench-201, and RDARTS search spaces. In DARTS, it reaches 97.42% top-1 on CIFAR-10 and 83.38% on CIFAR-100. In NAS-Bench-201, it discovers optimal or near-optimal architectures. In RDARTS, it achieves the best reported performance among the compared methods on two of four search spaces on CIFAR-10. We further analyze why MIDAS works, showing that patchwise attention improves discrimination among candidate operations and that the resulting input-specific parameter distributions are class-aware.
| Comments: | Published in Transactions on Machine Learning Research (TMLR) |
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Neural and Evolutionary Computing (cs.NE) |
| Cite as: | arXiv:2602.17700 [cs.LG] |
| (or arXiv:2602.17700v2 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2602.17700 arXiv-issued DOI via DataCite |
|
| Journal reference: | Transactions on Machine Learning Research (2026) |
Submission history
From: Konstanty Subbotko [view email]
[v1]
Fri, 6 Feb 2026 23:16:41 UTC (176 KB)
[v2]
Mon, 26 Oct 2026 22:16:52 UTC (174 KB)
来源：arXiv cs.AI · arxiv.org