AI Paper Atlas

Research map / Architectures and efficiency / Neural architecture search

Search Space Synthesis for Parametric Functions

Felix Laarmann, Andreas Pauly, Sebastian Buschjäger, Andrea Bommert, Jakob Rehof

ICML 2026 · Architectures and efficiency › Neural architecture search · approach: Differentiable architecture search (DARTS)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use a Weisfeiler-Lehman kernel on the DAMG string-diagram graph instead of on the combinatory term tree (ablation)GP surrogate · Global search spacemean R2: -0.394 -> 0.367Improved
use DAMG kernel instead of subtree kernel (ablation)GP surrogate · Global search spacemean NMSE: 0.837 -> 0.781Improved
use DAMG kernel instead of subtree kernel (ablation)GP surrogate · Refined search spacemean R2: 0.244 -> 0.431Improved
restrict structural variance (refined search space) with DAMG kernel (ablation)GP surrogate · DAMG kernelmean R2: 0.367 -> 0.431Improved
restrict structural variance (refined search space) with DAMG kernel (ablation)GP surrogate · DAMG kernelmean NMSE: 0.781 -> 0.874Hurt
use Bayesian optimization instead of random searchsearch over synthesized spaces · synthesized search spacesbest candidate foundImproved
add search-space refinements (ablation)Bayesian optimization · synthesized search spacesoptimization performanceMixed

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy