Aug 2026· Superconductors Science and Technology· 0 citations
TL;DR
A conceptual framework is proposed in which LLMs act as high-level reasoning and knowledge-integration layers that connect experimental data, simulation outputs, literature, and expert knowledge across modalities, while numerical computation and real-time protection remain the responsibility of validated physics-based models and specialised machine-learning algorithms.
Abstract
Applied superconductivity research spans materials science, physics, cryogenics, and large-scale device and system engineering, generating highly fragmented data, heterogeneous models, and disconnected domain-specific knowledge representations. This fragmentation limits systematic cross-disciplinary reasoning (across experimental, computational, and operational domains), slows innovation, and hinders the translation of experimental insights into deployable technologies. Compared with traditional analytical and data-driven methods, large language models (LLMs) exhibit complementary strengths in cross-domain knowledge integration, contextual reasoning, and multimodal information fusion. This perspective examines how these capabilities may be combined with superconducting-physics constraints, symbolic representations, and structured experimental and simulation data. We propose a conceptual framework in which LLMs act as high-level reasoning and knowledge-integration layers that connect experimental data, simulation outputs, literature, and expert knowledge across modalities, while numerical computation and real-time protection remain the responsibility of validated physics-based models and specialised machine-learning algorithms. Potential applications include LLM-assisted superconducting materials discovery, automated fault and quench diagnostics of experimental systems and superconducting devices, smart manufacturing and intelligent quality control, and technical documentation and support. We further discuss the practical limitations of these systems, including data quality, multimodal alignment, inference latency, uncertainty, hallucination, privacy, and the need for human oversight. The paper concludes by outlining a roadmap for the trustworthy integration of LLMs into superconductivity research and engineering.
This Perspective systematically discusses Physics-Grounded Materials AI (PhysMat AI) as a unifying perspective for integrating physical knowledge into materials intelligence through five complementary roles: physics as prior knowledge, descriptors, constraints, verifiers, and infrastructure.
Yuhang Wang, Qian Wang, Seong‐Hoon Jang et al.· Advanced Functional Material...· 0 citations
Large language models frequently violate fundamental scientific principles when generating technical content, undermining their reliability in scientific applications. We introduce Scientific Feasibility Control SFC, a graph-structured conformal prediction framework that provides statistical guarantees for scientific reasoning validity through progressive absolute-coherent-factuality validation. Our approach decomposes scientific reasoning into atomic absolute-coherent-factuality units requiring both individual correctness against physical laws and logical substantiation from preceding context, addressing the cascade effect where early scientific errors contaminate subsequent reasoning steps. Unlike independence-based methods that treat claims in isolation, SFC models logical dependencies as approximate deducibility graphs and operates through real-time validation with dynamic branching when scientific violations are detected, the system branches to alternative generation paths using verified context as foundation. We demonstrate SFC across established scientific reasoning benchmarks including PhyX multimodal physics, MATH, ScienceQA, and ARC Challenge, achieving 50.1 percent accuracy on PhyX physics reasoning, substantially outperforming recent reasoning models including DeepSeek-R1 49.8 percent and GPT-4 45.8 percent while providing 91.7 percent scientific validity with formal conformal coverage guarantees at alpha equals 0.10 confidence level and reducing scientific law violations by 73 percent across multiple model architectures.
Automation is transforming scientific discovery by enabling systematic exploration of complex hypotheses. Large language models (LLMs) perform well across diverse tasks and promise to accelerate research, but often struggle with logical structures. Here, we present a framework for biological discovery integrating LLM-based agents with laboratory automation, guided by logical scaffolds incorporating symbolic relational learning, structured vocabularies and experimental constraints. This integration improves coherence and reliability in automated workflows. We couple this AI-driven approach to automated cell-culture and metabolomics platforms, enabling integrated hypothesis validation and refinement, yielding a flexible discovery system. The system identified novel interactions in Saccharomyces cerevisiae, including glutamate-induced growth inhibition in spermine-treated cells and aminoadipate's partial rescue of formic-acid stress. All hypotheses, experiments and data are captured in a graph database employing controlled vocabularies. Existing ontologies are extended, and a novel representation of scientific hypotheses is presented using description logics. This work demonstrates the potential for a reliable machine-driven discovery process in systems biology.
Daniel Brunnsåker, Alexander H. Gower, Prajakta Naval et al.· Journal of the Royal Society...· 2 citations
A unified framework is introduced that distills intermediate representations that explicitly encode the physical modeling process and adopt a two-stage post-training strategy, where supervised fine-tuning establishes structured modeling, and reinforcement learning with rubric-based feedback improves the quality of the modeling process.
Ye Zhang, Xuehang Guo, Rui Pan et al.· 0 citations
Heterogeneous catalysis presents a distinct challenge for artificial intelligence (AI). Data sets are often small and inconsistently reported, catalyst representations are not standardized, and extracting fundamental knowledge requires integrating performance data, spectroscopic characterizations, and mechanistic models across multiple scales. Language offers a unifying representation across these modalities, making catalysis well suited for leveraging large language models (LLMs). By standardizing how catalytic data is represented, LLMs make dispersed experimental results more accessible to downstream statistical modeling. In this perspective, we focus our discussion around three opportunities where LLMs can significantly contribute to catalysis: (1) text to properties; (2) text to structure; and (3) text to mechanistic models. The discussion is followed by a perspective section on LLM-readiness of data, aligning LLM outputs with scientific correctness, and bridging lab-scale discovery to industrial deployment. Across each area, the most productive applications couple dispersed chemical knowledge with physics-grounded validation to produce verifiable hypotheses and actionable representations.
Shane S Michtavy, Sinhara M. H. D. Perera, Marc D. Porosoff· The journal of physical chem...· 0 citations
This tutorial provides a comprehensive, end-to-end view of LLM interpretability, transitioning from microscopic neural analysis to macroscopic application and deployment, and explores how these interpretability paradigms scale and inspire the design of frontier architectures, agentic systems, and thinking models.
Wei Zhang, Zhengfu He, Lucia Zhang et al.· Proceedings of the 32nd ACM...· 0 citations