Sharon Li · University of Wisconsin-Madison · PhdFit
Resume-aware faculty matching
Find professors who actually fit you
Upload your resume. PhdFit's six research agents compare your background with faculty profiles, recent publications, lab focus, and outreach opportunities, then rank professors with evidence you can review.
Resume-aware matchingEvidence from recent researchOutreach drafts included
Advisor Match Brief Resume uploaded
Resume signals5 found
NLPHCIFairnessGraph MLPyTorch
Six-agent analysisRunning
Axel · Candidate AnalystResume parsed
Nova · Professor ResearcherRecent papers
Echo · Match ExplainerEvidence linked
Atlas · Strategy AdvisorFit ranked
Lyra · Outreach WriterDraft queued
Juno · Meeting Prep CompilerNext step ready
SC
Dr. Sarah Chen
Stanford · NLP & interpretability
91
fit
Fairness + NLP projects overlap with her recent work.
Strong methods fit: model evaluation and interpretability.
Clear outreach angle for current trustworthy AI research.
Outreach angle
Ask how her lab is extending interpretability methods into fairness audits for real-world AI systems.
Sharon Li
· ProfessorVerified
University of Wisconsin-Madison · Computer Sciences
Yong Jae Lee is a Professor in the Department of Computer Sciences at the University of Wisconsin-Madison. His core research interests are in computer vision and machine learning, with a focus on creating robust AI systems that can learn to understand our multimodal world with minimal human supervision. Before joining UW-Madison in 2021, he spent one year as an AI Visiting Faculty at Cruise, and before that, six years as an Assistant and then Associate Professor at UC Davis. He received his Ph.D. from the University of Texas at Austin in 2012 advised by Kristen Grauman, and was a postdoc at Carnegie Mellon University (2012-2013) and UC Berkeley (2013-2014) advised by Alyosha Efros. He received his B.S. in Electrical Engineering from the University of Illinois at Urbana-Champaign in 2006. He is a recipient of numerous awards including the Army Research Office Young Investigator Program Award, the National Science Foundation CAREER Award, and the UW-Madison SACM Student Choice Professor of the Year Award. He and his collaborators have also received the Most Innovative Award at the COCO Object Detection Challenge ICCV 2019 and the Best Paper Award at BMVC 2020.
Ultralight Mg–Li alloys demonstrate significant potential for critical structural applications, yet their widespread adoption has been constrained by insufficient specific strength. Through multi-pass rotary swaging processing, we significantly enhance the mechanical properties of Mg–4Li alloy, achieving an exceptional specific strength of 214 kN•m/kg—far surpassing those of conventional binary Mg–Li alloys—alongside a yield strength of 277 MPa and a tensile strength of 343 MPa. This remarkable improvement stems from the formation of a refined nanograin structure. Comprehensive microstructural analysis identifies two primary refinement mechanisms: (1) dislocation-mediated grain subdivision and (2) twin-twin intersection-induced grain refinement, which collectively contribute to the development of nanocrystalline domains. The achieved grain refinement efficiency demonstrates rotary swaging's unique capability in processing hard-to-refine Mg alloys, overcoming the intrinsic limitations of conventional thermo-mechanical methods. This breakthrough suggests new possibilities for developing ultrahigh-strength Mg–Li systems. • Introduced a method of rotary swaging which suggesting new possibilities for developing ultrahigh-strength Mg–Li systems. •Achieved an exceptional specific strength of 214 kN m/kg, far surpassing conventional binary Mg–Li alloys. •Found the dislocation-mediated grain subdivision and twin-twin intersection-induced grain refinement mechanism.
The Journal of Physical Chemistry C · 2023 · 5 citations
Materials science
Chemical physics
Nanotechnology
The nucleation threshold of nanobubble formation around the surface of nanoparticles with small sizes is significantly high under the irradiation of incident light, which has limited their applications in the fields of optofluidics and biomedicine. To solve this problem, nanoparticles are assembled into an array to investigate the generation process of nanobubbles around them under the impact of the localized surface plasmon resonance (LSPR)-induced enhanced electric field (EEF) by using molecular dynamics (MD) simulations and continuum equations. The results of MD simulations indicate that even if the nanoparticle temperature is close to the melting point, it is still difficult to generate nanobubbles around the surface of an individual nanoparticle with a radius of 2 nm. However, when nanoparticles are assembled into an array, the progressive nucleation and generation process of nanobubbles around their surface can be observed after a thermal diffusion process lasting approximately 1050 ps. When the LSPR effect is excited, water molecules can be rapidly transformed from the liquid state to the gaseous state under the perturbation of an intensive EEF, which can further promote the generation process of nanobubbles. Moreover, using the thermal diffusion equation with a modification of the interfacial thermal resistance, a detailed analysis is also conducted for the phase change process of the water region surrounding nanoparticles under the effect of the LSPR-induced EEF. Compared with the results of MD simulations, it can be observed that the modified thermal diffusion equation can clearly demonstrate the initial nucleation of nanobubbles around the surface of a nanoparticle array but the size of the generated nanobubble is slightly smaller. On this basis, the generation process of nanobubbles can be flexibly manipulated by adjusting the gap distance of an array, the number of nanoparticles in an array, and the polarization direction of the incident light. The results could provide a new approach to investigate the nucleation and generation process of nanobubbles, which makes them more attractive in various emerging applications.
The conventional direct torque control (DTC) for permanent-magnet synchronous machines (PMSMs) uses hysteresis comparators to determine a single voltage vector per control period based on the torque and stator flux tracking errors. The saturation controller-based DTC (SDTC) is a duty cycle DTC strategy that uses nonlinear adaptive midpoint saturation controllers to determine the duration of two active voltage vectors and a zero voltage vector for each control period. A new voltage vector table and the outputs of the saturation controllers determine the duty cycle for each phase. This paper proposes a novel model predictive SDTC (MPSDTC) strategy for PMSMs with low sampling frequency, fast transient response, and low steady-state torque and flux ripples. The MPSDTC uses a computational model of the SDTC to predict the torque, stator flux linkage, current, etc. of the PMSM in the immediate next control period and possibly future control periods. The predicted values are then used to form a cost function that is minimized to find the optimized saturation controller outputs. The MPSDTC has better transient and steady state torque characteristic than the traditional field-orientated control (FOC) or the SDTC. 1
Computer Animation and Virtual Worlds · 2020 · 2 citations
Computer Science
Computer Science
Computational science
Abstract The accuracy of advection has a great influence on the visual effect of fluid simulation. Constrained interpolation profile (CIP) method has been an important advection scheme because of its third‐order accuracy and the fact that it only needs to be performed over a compact stencil, but extending it to high‐dimensional advection equations is not easy, because it involves complex calculations and large memory overheads, and is usually unstable. In this article, we propose a stable and efficient three‐dimensional (3D) CIP scheme which can maintain high accuracy but requires low computation and memory cost. We first construct an efficient two‐dimensional (2D) CIP scheme based on dimensional splitting and local Taylor expansions, and then propose an effective way to extend it for 3D applications without decreasing the computational accuracy or affecting the stability. The experimental results show the advantages of our method over the state‐of‐the‐art advection schemes.
Computer Animation and Virtual Worlds · 2020-09-30 · 3 citations
article
Abstract The advection step in grid‐based fluid simulation is prone to numerical dissipation, which results in loss of detail. How to improve the advection accuracy to preserve more fluid details is still challenging. On the other hand, a common way to enhance smoke details is to use vorticity confinement. However, most of the previous methods simply used a fine‐tuned scale factor ε to adjust the strength of the confinement force, which can only amplify existing vortex details and is easy to cause instability when ε is large. In this article, we proposed an adaptive vorticity confinement method, which does not suffer from the above problems, to compensate the vorticity loss during advection with little extra cost. The main idea is to first calculate a scale factor whose value depends on the vorticity loss during advection, and then use it to adaptively control the vorticity confinement force for vorticity compensation with high stability. The experiment results show the effectiveness and efficiency of our method.
SID Symposium Digest of Technical Papers · 2019-09-01
article1st authorCorresponding
The display is an essential part of our life, Micro‐LED (Micro‐Light Emitting Diode), as an emerging display technology, competes with OLED (Organic Light Emitting Diode) and QLED (Quantum‐dot Light Emitting Diode) to become the mainstream technology of the next generation of display technology in the display industry. Due to its advantages such as high resolution, high brightness, high response frequency, high stability, low power consumption, Micro‐LED has very big application prospect. GaN‐based HEMT (High Electron Mobility Transistor) also is the research hot point due to its high electron mobility. Due to they are the same material, it is possible to combine two of them. In this paper, a homogenous integration method of HEMT driver circuit of 2T1C (2 transistors and 1 capacitor) structure with micro LEDs is proposed and simulated by Silvaco software.
