Accelerating Quantum Materials Characterization: Hybrid Active Learning for Autonomous Spin Wave Spectroscopy

TL;DR

This paper introduces TAS-AI, a hybrid framework for autonomous spin-wave spectroscopy that separates detection, inference, and refinement tasks, improving efficiency and accuracy in quantum materials characterization.

Contribution

The paper presents TAS-AI, a novel hybrid active learning framework that explicitly separates detection, inference, and refinement in autonomous neutron spectroscopy, demonstrating improved performance over physics-informed methods.

Findings

Model-agnostic methods outperform physics-informed planning in initial detection tasks.

TAS-AI achieves decisive model discrimination with fewer measurements.

Motion-aware scheduling reduces wall-clock time by 32%.

Abstract

Autonomous neutron spectroscopy must solve three distinct tasks: detection (where is the signal?), inference (which Hamiltonian governs it?), and refinement (what are the parameters?). No single controller solves all three equally well. We present TAS-AI, a hybrid agnostic-to-physics-informed framework for autonomous triple-axis spin-wave spectroscopy that separates these tasks explicitly. In blind reconstruction benchmarks, model-agnostic methods such as random sampling, coarse grids, and Gaussian-process mappers reach a global error threshold more reliably and with fewer measurements than physics-informed planning, supporting the claim that discovery and inference are distinct tasks requiring distinct controllers. Once signal structure is localized, the physics-informed stage performs in-loop Hamiltonian discrimination and parameter refinement: in a controlled square-lattice test between nearest-neighbor-only and J1-J2 Hamiltonians, TAS-AI reaches a decisive AIC-derived evidence ratio (>100) in fewer than 10 measurements, while motion-aware scheduling cuts wall-clock time by 32% at a fixed measurement budget. We also identify a failure mode of posterior-weighted design, algorithmic myopia, in which the planner over-refines the current leading model while under-sampling low-intensity falsification probes. A constrained falsification channel sharply reduces time spent committed to the wrong model and accelerates correct model selection without modifying the Bayesian inference engine. In controlled two-model ablations, both a deterministic top-two max-disagreement rule and an LLM-based audit committee achieve this gain under identical constraints. We demonstrate the full workflow in silico using a high-fidelity digital twin and provide an open-source Python implementation.