About

Min Li is a professor at the University of Washington College of Education, with research interests centered on the development of children and youth, measurement and statistics, and quantitative research methods. His work emphasizes assessments that go beyond measuring learning outcomes to empower students and improve learning processes. His research aims to study and model how student learning can be accurately, adequately, and fairly assessed in both large-scale testing and classroom settings, especially in tech-rich environments, to provide rich and actionable assessment results for learners and educators. His approach combines cognitive sciences and psychometric modeling approaches within STEM disciplines, focusing on issues of validity and validation, including examining cognitive demands of science items, using natural language processing to detect testing bias, modeling student reasoning and problem-solving strategies, automating grading of written responses, and addressing measurement issues in instructional tasks. His recent projects include developing equitable assessments in computer science, extracting actionable results from diagnostic assessments, and creating authentic and fair assessments in STEM fields. His scholarly contributions include numerous publications on topics such as science assessment, student reasoning, and test validity.

Research topics

• Computer Science
• Artificial Intelligence
• Machine Learning
• Optoelectronics
• Materials science
• Data science
• Telecommunications
• Human–computer interaction
• Physics
• Optics

Selected publications

• On-the-fly closed-loop materials discovery via Bayesian active learning

  Nature Communications · 325 citations

  • Computer Science
  • Artificial Intelligence
  • Computer Science

  Abstract Active learning—the field of machine learning (ML) dedicated to optimal experiment design—has played a part in science as far back as the 18th century when Laplace used it to guide his discovery of celestial mechanics. In this work, we focus a closed-loop, active learning-driven autonomous system on another major challenge, the discovery of advanced materials against the exceedingly complex synthesis-processes-structure-property landscape. We demonstrate an autonomous materials discovery methodology for functional inorganic compounds which allow scientists to fail smarter, learn faster, and spend less resources in their studies, while simultaneously improving trust in scientific results and machine learning tools. This robot science enables science-over-the-network, reducing the economic impact of scientists being physically separated from their labs. The real-time closed-loop, autonomous system for materials exploration and optimization (CAMEO) is implemented at the synchrotron beamline to accelerate the interconnected tasks of phase mapping and property optimization, with each cycle taking seconds to minutes. We also demonstrate an embodiment of human-machine interaction, where human-in-the-loop is called to play a contributing role within each cycle. This work has resulted in the discovery of a novel epitaxial nanocomposite phase-change memory material.

  DOI

• Programmable phase-change metasurfaces on waveguides for multimode photonic convolutional neural network

  Nature Communications · 337 citations

  Senior authorCorresponding
  • Computer Science
  • Optoelectronics
  • Computer Science

  Abstract Neuromorphic photonics has recently emerged as a promising hardware accelerator, with significant potential speed and energy advantages over digital electronics for machine learning algorithms, such as neural networks of various types. Integrated photonic networks are particularly powerful in performing analog computing of matrix-vector multiplication (MVM) as they afford unparalleled speed and bandwidth density for data transmission. Incorporating nonvolatile phase-change materials in integrated photonic devices enables indispensable programming and in-memory computing capabilities for on-chip optical computing. Here, we demonstrate a multimode photonic computing core consisting of an array of programable mode converters based on on-waveguide metasurfaces made of phase-change materials. The programmable converters utilize the refractive index change of the phase-change material Ge2Sb2Te5 during phase transition to control the waveguide spatial modes with a very high precision of up to 64 levels in modal contrast. This contrast is used to represent the matrix elements, with 6-bit resolution and both positive and negative values, to perform MVM computation in neural network algorithms. We demonstrate a prototypical optical convolutional neural network that can perform image processing and recognition tasks with high accuracy. With a broad operation bandwidth and a compact device footprint, the demonstrated multimode photonic core is promising toward large-scale photonic neural networks with ultrahigh computation throughputs.

  DOI

• VL-Anodiff: Vision-Language Guided Diffusion for Few-Shot Industrial Anomaly Synthesis

  2026-04-21

  article1st authorCorresponding

  Acquiring large-scale annotated defect data is costly and often impractical in industrial scenarios, making few-shot anomaly synthesis a promising solution to enhance anomaly detection. While existing approaches based on handcrafted augmentations or generative models have shown encouraging results, they either struggle with semantic authenticity and controllability or rely on manually defined bounding boxes to guide anomaly synthesis. To address these limitations, we introduce VL-AnoDiff, a vision-language guided diffusion framework for few-shot industrial anomaly synthesis. Firstly, we propose Semantic Anchor Regularization (SAR), which aligns learnable embeddings with semantic anchors derived from vision-language models, thereby ensuring semantically faithful and realistic anomaly content. Moreover, a lightweight Semantically Aligned Mask Synthesis (SAMS) mechanism, which incorporates semantic priors, is employed to generate diverse and spatially meaningful masks, thereby eliminating the need for bounding box supervision or extensive training. Extensive experiments on multiple benchmarks demonstrate that VL-AnoDiff produces high-fidelity and controllable anomalies, and significantly boosts downstream anomaly inspection performance under few-shot settings.

  PublisherDOI

• Stability Analysis for Delayed Neural Networks Based on Delay-Derivative-Dependent Approaches

  IEEE Access · 2026-01-01

  articleOpen access

  The time delay and its derivative are commonly constrained by upper and lower bounds for the stability analysis of delayed neural networks. To make fuller use of the delay-related information, in the current paper, some essential methods, such as the integral inequality, the positive definite activation-function-based cross terms and delay-product Lyapunov-Krasovskii functional are refined by extra introducing the information of delay derivative. That is, these refined methods utilize not only delay-dependent terms but also delay-derivative-dependent ones, and therefore they are less conservative. By means of these refined methods, a delay-derivative-dependent stability criterion, introducing less conservatism, is established for delayed neural networks, using two examples to verify the asvatages.

  PublisherDOI

• Physics-informed solar modulation attention based dual-branch Transformer for ultra-short-term photovoltaic power forecasting with operational benefits analysis

  SSRN Electronic Journal · 2026-01-01

  preprintOpen access
  PublisherDOI

• Experimental study of acoustic loss at microwave frequencies in thin-film lithium niobate

  Applied Physics Letters · 2026-04-27

  articleSenior author

  Thin-film lithium niobate (TFLN) has emerged as a versatile platform for phononic and photonic devices with applications ranging from classical signal processing to quantum technologies. However, acoustic loss fundamentally limits the performance of acoustic devices on TFLN platforms, yet its physical origin remains insufficiently understood. Here, we systematically investigate acoustic propagation loss in various TFLN platforms, including lithium niobate on insulator (LNOI), lithium niobate on sapphire, suspended lithium niobate (LN) thin films, and bulk LN at gigahertz frequencies over temperatures ranging from 4 K to above room temperature. Using a delay-line method, we extract frequency- and temperature-dependent losses for Rayleigh, shear-horizontal, and Lamb modes. We observe an anomalous non-monotonic temperature dependence in LNOI that closely resembles acoustic loss in amorphous materials, suggesting a loss channel associated with the buried oxide layer at low temperatures. At elevated temperatures, the loss converges to the Akhiezer damping governed by phonon–phonon interactions. High-resolution electron microscopy further reveals nanoscale interfacial crystal impurities that may contribute to the increased acoustic loss in TFLN platforms relative to bulk LN. These results elucidate the acoustic loss mechanisms in TFLN and provide guidelines for designing low-loss acoustic devices.

  PublisherDOI

• Optical multi-beam steering and communication using integrated acousto-optics arrays

  Nature Communications · 2025-05-15 · 13 citations

  articleOpen accessSenior author

  Optical beam steering enables optical sensing, imaging, and long-range communication over free space. Despite the inherent speed of light, advanced applications increasingly require simultaneous steering of multiple, independently controlled beams, to enhance imaging throughput, boost communication bandwidth, and control qubit arrays for scalable quantum computing. However, precise multi-beam steering and control remain a significant challenge with current solid-state beam steering technologies, driving the need for integrated and scalable multi-beam steering solutions. Here, we report a scalable multi-beam steering system comprising an array of integrated acousto-optic beam steering channels on a thin-film lithium niobate platform. Each channel generates tens of individually controllable beams at 780 nm with sub-microsecond switching time by exciting acoustic waves using multi-tone microwave signals. We demonstrate the system’s unique capabilities through multiple-input, multiple-output free-space communications, simultaneously transmitting to multiple receivers at megabits/sec data rates. This technology is readily scalable to steer hundreds of optical beams from a compact chip, potentially advancing many areas of optical technologies and enabling novel applications. Solid-state optical beam steering is crucial for a wide array of optical technologies. Here, the authors present a chip-scale multi-beam steering system using an acousto-optic array. Each beam is individually controllable, enabling simultaneous transmission of multiple data streams to multiple receivers for advanced free-space optical communication.

  PublisherOA PDFDOI

• Displacement Damage Effects of Low-Energy Gallium Ion Irradiation on Single-Walled Carbon Nanotube Field-Effect Transistors

  Chinese Physics B · 2025-09-30

  articleOpen access

  Abstract The displacement damage (DD) effects induced by low-energy gallium ions (Ga+) on single-walled carbon nanotube field effect transistors (SWCNT FETs) are investigated in this study. Exposure to 5 keV Ga + irradiation resulted in significant changes in the Raman spectra and electrical properties of the devices. The key finding reveals a significant heavy ion energy-dependence in displacement damage (DD): the displacement damage dose (D d ) induced by 5 keV Ga + irradiation is nearly three orders of magnitude higher than that induced by 2225 MeV xenon ions (Xe + ). By integrating Raman spectroscopy, electrical characterization, and TRIM simulations, we demonstrate that low-energy heavy ions deposit substantially more energy via non-ionizing energy loss (NIEL) processes within the SWCNT and gate oxide layers compared to high-energy ions. This enhanced energy deposition generates increased atomic displacements and vacancies, which significantly degrade both the conductivity of the SWCNT channel and the insulating properties of the gate oxide. These findings provide critical insights into the impact of low-energy ion irradiation on SWCNTs and contribute to a deeper understanding of SWCNT FET behavior in radiation environments.

  PublisherDOI

• Low-noise and widely-tunable THz signal generation from a PDH-locked Brillouin laser system

  2025-12-08

  article

  This work proposes a photonic system for generating widely tunable THz signals with low phase noise. By Pound- Drever-Hall (PDH) locking two free-running CW lasers to the same high Q optical fiber resonator, the system achieves widely tunable THz signal generation. Exploiting the linewidth compression effect of stimulated Brillouin scattering in the resonator, the synthesized THz signals exhibit low phase noise. In the experimental demonstration, THz signals were generated across 90–300 GHz. Phase noise measurements in the 90–140 GHz band reached below –105 dBc/Hz at 10 kHz offset.

  PublisherDOI

• Stability Analysis of a Time-Delay Load Frequency Control System via an Improved Matrix-Separation-Based Inequality

  Energies · 2025-10-25

  articleOpen accessSenior author

  This study focuses on the stability of time-delay load frequency control (LFC) systems. Based on the Lyapunov–Krasovskii (L–K) functional method, a stability criterion with less conservatism and lower computational complexity is proposed. Unlike recent methods that decrease conservatism through enhancing the complexity of L–K functional, only the double integral is augmented in this paper. To estimate the L–K functional derivatives more precisely, an improved matrix-separation-based inequality is proposed, which introduces some delay-derivative-dependent matrices rather than the high-dimensional free matrices. By applying the augmented L–K functional and the improved matrix-separation-based inequality, the stability criterion is established. Case analysis demonstrates that the new stability criterion has less conservatism and lower computational complexity, thereby validating the correctness of the method presented.

  PublisherOA PDFDOI

Recent grants

• CAREER: Integration of 2D materials for broadband silicon photonics

  NSF · $400k · 2014–2019

• CAREER: Integration of 2D materials for broadband silicon photonics

  NSF · $90k · 2018–2020

• Collaborative Research: Nanofabricated multisite opto-electrical probes for optogenetic stimulation and neuronal recording applications

  NSF · $311k · 2013–2017

• Collaborative Research: Quantum acoustics for optomechanical transduction and entanglement of solid-state spin qubits

  NSF · $468k · 2020–2023

• Integrated surface acousto-optical devices on piezoelectric aluminum nitride thin films

  NSF · $359k · 2013–2017

Frequent coauthors

• Ying Zhang

  Jining Medical University

  372 shared

• Jing Yang

  Southern Medical University

  72 shared

• Huan Zhang

  University of Chinese Academy of Sciences

  66 shared

• Tao Chen

  Xi'an Polytechnic University

  63 shared

• Yan Sun

  Peking University

  54 shared

• Huan Li

  52 shared

• Jian Zhang

  Zhejiang Academy of Forestry

  52 shared

• Jiayi Chen

  48 shared

Labs

• Measurement & StatisticsPI

Education

• Ph.D., Applied Physics

  California Institute of Technology

  2007