analiza aiqbench
This research introduces AutoMat, an agentic framework that solves the ill-posed inverse problem of reconstructing crystal structures from noisy STEM images by integrating physics-guided modules with closed-loop verification. By employing a failure-aware controller capable of hypothesis search and rollback-and-retry mechanisms, the system outperforms standard feed-forward models and specialized domain tools in generating physically valid atomic-scale simulations. The work establishes a significant pathway for automating materials science workflows through advanced AI agent orchestration rather than simple image processing.
AutoMat: Enabling Automated Crystal Structure Reconstruction from Microscopy via Agentic Tool Use
arXiv:2505.12650v2 Announce Type: replace-cross Abstract: Reconstructing atomistic crystal structures from a single noisy STEM projection is an ill-posed inverse problem: multiple lattices can explain similar contrast, and purely feed-forward models cannot verify physical validity. We present AutoMat, a failure-aware agentic controller that performs inference-time hypothesis search with closed-loop verification to convert Scanning Transmission Electron Microscopy (STEM) images into simulation-ready crystal structures and downstream properties. AutoMat composes perception and physics modules---pattern-adaptive denoising, physics-guided template retrieval as a state-dependent auxiliary branch, symmetry-constrained atomic reconstruction, and MLIP-based relaxation/validation---and triggers rollback-and-retry when verification fails. For systematic evaluation, we introduce STEM2Mat-Bench, a benchmark dataset containing 450+ annotated samples. Performance is assessed using lattice root-mean-square deviation (RMSD), formation energy mean absolute error (MAE), and structure matching accuracy. Results demonstrate that AutoMat outperforms existing approaches including SOTA models, specialized domain tools, and closed-source multimodal large models. This work establishes a direct pathway from microscopic characterization to atomic-scale modeling, addressing a fundamental challenge in materials science.
Abstract:Reconstructing atomistic crystal structures from a single noisy STEM projection is an ill-posed inverse problem: multiple lattices can explain similar contrast, and purely feed-forward models cannot verify physical validity. We present AutoMat, a failure-aware agentic controller that performs inference-time hypothesis search with closed-loop verification to convert Scanning Transmission Electron Microscopy (STEM) images into simulation-ready crystal structures and downstream properties. AutoMat composes perception and physics modules---pattern-adaptive denoising, physics-guided template retrieval as a state-dependent auxiliary branch, symmetry-constrained atomic reconstruction, and MLIP-based relaxation/validation---and triggers rollback-and-retry when verification fails. For systematic evaluation, we introduce STEM2Mat-Bench, a benchmark dataset containing 450+ annotated samples. Performance is assessed using lattice root-mean-square deviation (RMSD), formation energy mean absolute error (MAE), and structure matching accuracy. Results demonstrate that AutoMat outperforms existing approaches including SOTA models, specialized domain tools, and closed-source multimodal large models. This work establishes a direct pathway from microscopic characterization to atomic-scale modeling, addressing a fundamental challenge in materials science.
Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
arXiv:2505.12650 [cs.CV]
(or arXiv:2505.12650v2 [cs.CV] for this version)
Text integral sub licența deschisă a publicației · 170 cuvinte
Subiecte conexe
News
Quantum Computing Report · 25 iul. 2026
Researchers Automate Creation of Entangled Photonic Circuits
Quantum Zeitgeist · 17 sept. 2026
The Machines that Make the Machines
NVIDIA · 7 oct. 2026
Synthesis Superintelligence: from Semiconductors to Superconductors — Periodic Labs’ Liam Fedus and Ekin Dogus Cubuk
Latent Space · 8 oct. 2026