Research map / Architectures and efficiency / Neural architecture search
Cool Graphs: Active Property Search Towards Quantum Nano-Refrigerators
Sofia Blyufer, Amir Kleiner, Uri Peskin
Read the paperCheck your idea against 137,217 papers
AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Proposes an active property search approach to efficiently find graph/nanodevice configurations that act as quantum refrigerators, avoiding expensive quantum-chemical computations for each candidate.
Problem
Finding nanodevice configurations that function as quantum refrigerators is a needle-in-a-haystack search: the many-body state space grows exponentially with device size and each candidate needs an expensive quantum-chemical calculation, with only 2-9% of random configurations showing the desired be
Approach
Treats configuration discovery as an active search problem over graphs, using sequential selection of candidates to find rare refrigerating configurations with far fewer expensive quantum-chemical evaluations than random sampling. Details beyond the abstract are not available.
Key results
- Random sampling of material parameters yields only 2-9% of configurations exhibiting the desired refrigeration behavior, motivating the active search approach.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use uncertainty-driven (MC dropout) acquisition instead of random acquisition | GVAE active learning loop · 11 topologies, 16,000 oracle calls | positives found: 670 -> 4477 | Improved |
| use deep-ensemble uncertainty instead of MC dropout | active learning loop · 16,000 oracle calls | positives found: 4477 -> 4925 | Improved |
| use uncertainty-driven acquisition instead of GP Bayesian optimization | search strategy · 16,000 oracle calls | positives found: 1.46-1.53x more positives | Improved |
| retrain classifier on pool from uncertainty acquisition instead of random acquisition | GVAE · extrapolation test | ROC-AUC: 0.911 -> 0.934 | Improved |
| retrain classifier on pool from GP-BO acquisition (more positives) instead of random | GVAE · interpolation test | ROC-AUC: 0.932 -> 0.924 / 0.923 | No clear effect |
| add variational (KL) term vs plain autoencoder (ablation) | graph autoencoder · interpolation test | PR-AUC: 0.354 -> 0.409 | Improved |
| add auxiliary decoders (reconstruction) to discriminative classifiers (ablation) | GAT/GCN classifiers · interpolation test | PR-AUC: 0.07-0.12 higher | Improved |
| remove the KL term entirely (ablation) | GVAE · interpolation/extrapolation test | PR-AUC: -0.017 (0.010) | No clear effect |
| multiply KL weight by 10 (ablation) | GVAE · interpolation test | PR-AUC: -0.108 | Hurt |
| weaken reconstruction weight tenfold (ablation) | GVAE · interpolation test | PR-AUC: -0.083 | Hurt |
| raise latent dimension from 32 to 128 (ablation) | GVAE · interpolation test | PR-AUC: -0.138 | Hurt |
| inject random selection every eleventh active-learning cycle (ablation) | uncertainty acquisition · per-cycle positive rate | positive rate: 0.280 -> 0.281 | No clear effect |
Limitations
- Full text unavailable; limitations not determinable from the abstract.
Experimental setup
Baselines: random sampling
Closest papers by meaning
- Learning the Electronic Hamiltonian of Large Atomic Structures (ICML 2025 · similarity 0.86)
- TITAN: A Trajectory-Informed Technique for Adaptive Parameter Freezing in Large-Scale VQE (NeurIPS 2025 · similarity 0.86)
- Learning Smooth and Expressive Interatomic Potentials for Physical Property Prediction (ICML 2025 · similarity 0.84)
- Scalable Quantum-Inspired Optimization Through Dynamic Qubit Compression (AAAI 2025 · similarity 0.84)
- QCircuitBench: A Large-Scale Dataset for Benchmarking Quantum Algorithm Design (NeurIPS 2025 · similarity 0.84)
- TensorRL-QAS: Reinforcement learning with tensor networks for improved quantum architecture search (NeurIPS 2025 · similarity 0.84)
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.
