Find Academic Supervisors

Antoine BosselutView profile

Assistant Professor

• Natural Language Processing
• Artificial Intelligence
• Machine Learning
• +5 more

Antoine Bosselut is a Tenure Track Assistant Professor at École Polytechnique Fédérale de Lausanne (EPFL) in the School of Computer and Communication Sciences, where he leads the EPFL NLP group. His research focuses on developing AI reasoning agents that can model, represent, and reason about human and world knowledge, with applications in health, education, and global fairness. His research interests span multiple critical areas in modern AI: LLM Representations of Knowledge: Understanding what language models know and how they represent that knowledge internally Reasoning Algorithms: Developing methods to improve LLMs' reasoning capabilities through symbolic systems, neuroscience, and cognitive science Large-scale AI Development: Creating open-source foundation models with multilingual capabilities AI Democratization: Ensuring equitable access to AI technologies across different cultural and regulatory contexts His recent publications reveal a strong focus on the intersection of language models and cognitive science, with multiple studies examining how LLMs align with human cognition. He's also deeply engaged in practical applications of NLP technology, particularly in education and multilingual settings, as evidenced by his work on evaluating AI's impact on higher education and developing culturally-aware language models. His scientific achievements have been recognized with several prestigious awards: Outstanding Paper Award at NAACL 2025 ELLIS Scholar designation in 2024 Outstanding Paper Award at ACL 2023 Inclusion in Forbes 30 Under 30 list for Science & Healthcare in 2021 Bosselut actively mentors numerous PhD students across diverse NLP research areas and has secured significant recognition for his work in both academic and media circles. His lab has been featured in major publications including Communications of the ACM, The Atlantic, and Quanta Magazine, highlighting the societal impact of his research on commonsense reasoning and AI capabilities.

Oscar Randal-WilliamsView profile

Professor

• Algebraic Topology
• Geometric Topology
• Mapping Class Groups
• +8 more

Oscar Randal-Williams is the Sadleirian Professor of Pure Mathematics at the University of Cambridge, where he is affiliated with the Faculty of Mathematics and the Department of Pure Mathematics and Mathematical Statistics (DPMMS). His work is centered in the Differential Geometry & Topology research group, where he contributes to advancing knowledge in geometric and algebraic topology. Professor Randal-Williams specializes in Algebraic and Geometric Topology, with particular expertise in mapping class groups, moduli spaces, cobordism categories, spaces of manifolds, surgery theory, configuration spaces, characteristic classes, and K-theory. His research explores the deep connections between homotopy theory and geometric structures, with applications across various mathematical domains. His work often bridges abstract algebraic structures with concrete geometric problems, creating new frameworks for understanding topological phenomena. An analysis of Professor Randal-Williams' recent publications reveals a strong focus on homological stability phenomena, mapping class groups of high-dimensional manifolds, and the interplay between algebraic structures and geometric topology. His work frequently involves E ∞ -algebras, general linear groups, and the topology of diffeomorphism groups. A significant portion of his research investigates the structure of moduli spaces of manifolds and their connections to algebraic K-theory, with recent work extending to applications in mathematical physics through studies of symmetries in quantum field theories. Professor Randal-Williams maintains active collaborations with leading mathematicians worldwide, including Søren Galatius, Alexander Kupers, and Jeremy Miller, among others. His research has been published in top-tier mathematical journals including the Annals of Mathematics, Inventiones Mathematicae, and the Journal of the American Mathematical Society.

Arman CohanView profile

Assistant Professor

• Natural Language Processing
• Machine Learning
• Artificial Intelligence
• +3 more

Arman Cohan is an Assistant Professor in the Department of Computer Science at Yale University, where he leads the Yale NLP Lab since its founding in January 2023. His research spans natural language processing and machine learning with emphasis on language modeling, representation learning, retrieval systems, and specialized domain applications including scientific discovery and AI for science. His primary research interests include: Natural Language Processing Machine Learning Large Language Models Information Retrieval AI for Science Scientific Problem-Solving Recent publications (2025) demonstrate intense focus on evaluating and advancing LLM capabilities across multimodal reasoning, scientific claim verification, financial domain applications, and biological modeling. The lab consistently produces high-impact work accepted at top-tier conferences including ACL, EMNLP, and ICLR, with 11 papers at ACL 2025 alone. Scientific awards include: Best Paper Award at AI4Research Workshop (IJCAI 2024) Outstanding Paper Award at EACL 2023 Best Paper Award at ACL 2024 for Olmo language model research Professor Cohan actively advises PhD students including Kaili Liu, Jacob Dunefsky, Alan Li, Yilun Zhao, and has graduated researchers such as Linyong Nan (now at Zoom) and Ansong Ni (now at Meta). The lab maintains strong industry partnerships and receives substantial research funding as evidenced by its prolific output and conference presence. The Yale NLP Lab hosts the annual New England NLP Workshop and regularly features speakers from leading institutions including Meta AI, Allen Institute for AI, and DeepMind, fostering a collaborative environment for advancing NLP research.

Adriana KovashkaView profile

Associate Professor

• Computer Vision
• Machine Learning
• Natural Language Processing
• +7 more

Adriana Kovashka is an Associate Professor at the University of Pittsburgh , affiliated with the School of Computing and Information and serving as Department Chair . Her academic journey began with BA degrees in Computer Science and Media Studies from Pomona College (2008) and a PhD in Computer Science from The University of Texas at Austin (2014). Joined Pitt’s faculty in January 2015 NSF CAREER awardee (2021) Google Faculty Research Award recipient Dr. Kovashka’s research spans Computer Vision , Machine Learning , and Natural Language Processing , focusing on visual rhetoric, weak multimodal supervision, and domain adaptation. She pioneered techniques for analyzing political imagery, developing robust object detection frameworks, and exploring the intersection of visual and textual persuasion through large-scale annotated datasets. Her recent work emphasizes geographic diversity in vision-language systems, audio-visual fusion for domain generalization, and shape-texture bias mitigation in CNNs. Key publications include groundbreaking studies on symbolic reasoning, multimodal dialogue systems, and ethical AI applications in education. Scientific honors include: NSF CRII Award (2016) NSF CAREER Award (2021) Pitt CRDF Award (2016, 2018) Best Paper at ECV Workshop (2021) Google Faculty Research Award (2016, 2018) Dr. Kovashka actively mentors students in multimodal learning projects and collaborates with interdisciplinary teams on NSF-funded initiatives. She co-organizes workshops like the first CVPR workshop on advertisement understanding and leads research groups exploring human-AI co-learning systems.

Robin JiaView profile

Assistant Professor

• Natural Language Processing (NLP)
• Deep Learning
• Artificial Intelligence
• +7 more

Robin Jia is an Assistant Professor in the Thomas Lord Department of Computer Science at the University of Southern California (USC) , where he leads the AI, Language, Learning, Generalization, and Robustness (Allegro) Lab . His research focuses on enhancing the reliability and robustness of large language models (LLMs) through mechanistic understanding, benchmarking under distribution shifts, and neurosymbolic integration. Key affiliations include collaborations with the USC Keck School of Medicine and contributions to legal frameworks like the EU's Digital Services Act. Robin's research spans multiple domains, including: Scientific analysis of LLM capabilities in in-context learning , data memorization , and numerical reasoning Advancements in robust NLP systems , emphasizing uncertainty estimation and calibration Development of methods combining LLMs with symbolic solvers for complex reasoning tasks Interdisciplinary applications in medicine and law , such as privacy-preserving synthetic data generation and medical misconception evaluation . His recent publications (2024-2025) address Fourier-based numerical embeddings (NeurIPS), neurosymbolic planning (NAACL), and multimodal benchmarking (COLM), with a strong emphasis on privacy , fairness , and transparency . Scientific awards include the Google Research Scholar Award (2023) , SoCalNLP Symposium Best Paper Awards , and ACL/EMNLP outstanding papers . He advises PhD students like Johnny Wei and Ameya Godbole, and has secured grants from the NSF , USC-Capital One , and USC-Amazon .

Greg DurrettView profile

Associate Professor

• Natural Language Processing
• Machine Learning
• Artificial Intelligence
• +7 more

Greg Durrett is an Associate Professor in the Department of Computer Science at University of Texas at Austin, leading the TAUR Lab (Text Analysis, Understanding, and Reasoning ). His research focuses on advancing Large Language Models (LLMs) for knowledge-intensive tasks in medical information processing scientific discovery legal reasoning . He received his B.S. in Computer Science and Mathematics from MIT (2010) and Ph.D. in Computer Science from UC Berkeley (2016). His work develops techniques to train LLMs with new capabilities augment models for reliability assess model outputs improve reasoning frameworks . His 15 most recent publications (2021-2025) span knowledge propagation in LLMs chain-of-thought reasoning code generation benchmarks multi-modal reasoning fact verification discourse analysis . Scientific honors include NSF CAREER Award (2024) NSF grants (2018, 2024) Bloomberg Data Science Grant (2017) Facebook Fellowship (2014) Best Paper Finalist (EMNLP 2013) . Teaching: CS388: Natural Language Processing (graduate) CS371N: NLP (undergraduate) High school NLP module .

Farzad MashayekView profile

Professor

• Turbulent Reacting Flow
• Plasma Flow
• Electrostatic Atomization
• +9 more

Farzad Mashayek is a Professor and Department Head of Aerospace and Mechanical Engineering at the University of Arizona, College of Engineering. He is a member of the Graduate Faculty and leads the Computational Multiphase Transport Laboratory. His research integrates high-fidelity simulations, machine learning, and experimental validation across diverse domains in fluid dynamics and energy systems. Educational Background: PhD in Mechanical Engineering, State University of New York at Buffalo, Buffalo, NY MS in Mechanical Engineering, Sharif University of Technology, Tehran, Iran BS in Mechanical Engineering, Sharif University of Technology, Tehran, Iran His research interests include turbulent reacting flows, plasma dynamics, electrostatic atomization, solid-ion and lithium batteries, computational fluid dynamics, and machine learning applications in engineering. He employs high-order spectral element methods, phase-field modeling, and deep neural networks to study complex multiphysics phenomena such as drop impact, battery degradation, and turbulence modeling. The recent publications reflect a strong trend toward integrating machine learning with multiphysics simulations, particularly in battery safety (thermal runaway prediction), materials characterization (STEM image analysis), and fluid dynamics (modal analysis of turbulence). His work often involves collaboration with experimental groups to validate models, especially in dental aerosol suppression and electrohydrodynamics. Scientific Awards: Sustained Service Award, American Institute of Aeronautics and Astronautics (AIAA), Spring 2022 Best Presentation Award, The 20th International Conference on Computational Mathematics, Parallel and Distributed Computing, Summer I 2018 Dr. Mashayek has secured funding from NSF (GOALI program) for controlled coating via charged droplet deposition. He advises graduate students and postdoctoral researchers in computational mechanics and energy systems, fostering interdisciplinary research. He has contributed to engineering education, particularly during the pandemic, with active learning strategies in online instruction. He leads a dynamic research team focused on advancing simulation tools and applying them to real-world challenges in energy, manufacturing, and public health.

Johnny GuzmánView profile

Professor

• Numerical Analysis
• Partial Differential Equations
• Scientific Computing
• +5 more

Johnny Guzmán is a Professor of Applied Mathematics at Brown University, specializing in numerical analysis of partial differential equations and scientific computing. He holds a Ph.D. in Applied Mathematics from Cornell University (2005) and a B.S. in Mathematics from California State University, Long Beach (1999). His research focuses on numerical methods for PDEs, including discontinuous Galerkin methods, mixed finite element methods, and fluid-structure interaction problems. Key contributions include work on hybridizable and mixed finite element methods, discontinuous Galerkin discretizations, and stability analysis of numerical schemes. He has been funded by multiple NSF grants, including a Postdoctoral Fellowship (2005–2008) and awards totaling over $1M in research support. Notable recognitions include the Comfort and Urry Family Fund Prize (2013). Guzmán collaborates with institutions globally and serves on editorial boards for journals like Journal of Numerical Mathematics and Calcolo . His teaching spans computational linear algebra, numerical methods for differential equations, and finite element analysis.

• Computer Vision
• Self-Supervised Learning
• Computational Photography
• +7 more

Alexei A. Efros is a Professor in the Department of Electrical Engineering and Computer Sciences (EECS) at UC Berkeley, where he holds the Howard Friesen Professorship and is affiliated with the Berkeley Artificial Intelligence Research (BAIR) Lab. He previously served on the faculty at the Robotics Institute of Carnegie Mellon University (CMU) and completed a postdoctoral fellowship at the University of Oxford. His research spans data-driven computer vision, self-supervised learning, computational photography, and applications to computer graphics and robotics. His research interests include: Data-Driven Computer Vision Self-Supervised and Unsupervised Learning Generative Models and Image Synthesis Visual Representation Learning Applications in Robotics and Human-Computer Interaction Intersections with Human Vision and the Humanities The recent publications highlight a strong trend toward self-supervised learning, visual reasoning, and generative modeling, particularly diffusion models and 3D scene understanding. His work increasingly bridges computer vision with language, robotics, and cognitive science, emphasizing interpretability and real-world applicability. There is a clear focus on leveraging unlabeled data and developing methods for robust, generalizable AI systems. His scientific awards and recognitions include: Berkeley Fellowship Google Fellowship Soros Fellowship NSF Fellowship SIGGRAPH Outstanding Doctoral Dissertation Award Facebook Fellowship Adobe Fellowship CMU School of Computer Science Distinguished Dissertation Award ACM Doctoral Dissertation Honorable Mention Alexei Efros has advised numerous PhD students and postdocs, many of whom have gone on to faculty positions at top institutions including CMU, Stanford, MIT, Columbia, NYU, and Georgia Tech. His lab has received research funding from major tech companies and federal agencies, though specific grants are not detailed in the text. He teaches core computer vision and machine learning courses at both undergraduate and graduate levels at UC Berkeley. His research group is highly active, with ongoing projects in 3D perception, generative modeling, and vision-language systems. He leads a vibrant research lab at UC Berkeley, part of the BAIR consortium, collaborating with leading researchers such as Jitendra Malik, Trevor Darrell, Pieter Abbeel, and Angjoo Kanazawa. His lab fosters strong interdisciplinary connections with institutions worldwide, including Oxford, INRIA, and École Normale Supérieure.

Jiatao GuView profile

Assistant Professor

• Machine Learning
• Generative Models
• 3D Generation
• +3 more

Jiatao Gu is an Assistant Professor in the Department of Computer and Information Science (CIS) at the University of Pennsylvania, with a part-time role as Staff Research Scientist at Apple (MLR). He holds a Ph.D. in Electrical and Electronic Engineering from the University of Hong Kong (2018) and a B.Eng. in Electronic Engineering from Tsinghua University (2014). His research focuses on generative machine learning and AI agent interaction with the physical world, emphasizing multi-modal systems spanning language, images, videos, and 3D. Key themes include efficient modeling , flexible architecture design , and scalable decision-making frameworks . 2025: ICLR paper on DART framework 2024: TMLR work on GFlowNet alignment 2023: NeurIPS research on diffusion stability 2022: ACL papers on speech translation Recent publications explore diffusion models for text-to-image synthesis, 3D reconstruction, and efficient sampling techniques. His work addresses fundamental challenges in attention mechanisms, entropy collapse, and multi-stage distillation while advancing non-autoregressive translation and vision-language reasoning . Prospective students can apply through his recruitment process at UPenn. Prior affiliations include Meta AI (FAIR Labs) and academic collaborations with institutions like New York University's CILVR Lab.

Jiaxuan YouView profile

Assistant Professor

• Graph Neural Networks
• Relational Data
• Machine Learning Systems
• +9 more

Jiaxuan You is an Assistant Professor at the Siebel School of Computing and Data Science, University of Illinois at Urbana-Champaign, leading the U Lab focused on achieving Artificial General Intelligence (AGI) in digital environments. His research spans graph neural networks (GNNs), relational data, foundation models, and machine learning systems. PhD and MS in Computer Science from Stanford University (2021) Developed GraphGym and PyTorch Geometric (PyG) for graph learning Core member at Kumo AI (2021-2023) His research explores: Graph-enhanced LLMs: Integrating relational structures into foundation models AGI Development: Self-optimizing AI agents and tool utilization ML Systems: Scalable architectures and redundancy-free computation Interdisciplinary Applications: Financial networks, crop yield prediction, and metro systems Recent publications focus on temporal reasoning, multi-agent dynamics, and hybrid architectures for LLMs. He actively develops open-source tools like DBGYM and GraphRouter. Scientific recognition includes: JPMC PhD Fellowship Baidu Scholarship Best Student Paper at AAAI 2017 World Bank Big Data Innovation Challenge winner He mentors PhD and intern students, emphasizing machine learning systems expertise. His lab collaborates on AGI workshops (e.g., ICLR 2024) and industry projects.

• Data Mining
• Matrix Factorizations
• Tensor Factorizations
• +2 more

Evrim Acar Ataman is a Research Professor and Chief Research Scientist at Simula Metropolitan, where she serves as Head of the Department of Data Science and Knowledge Discovery. Her research focuses on advanced data mining techniques for complex, multi-modal datasets across biomedical and network domains. Her primary research interests include Data Mining , Matrix and Tensor Factorizations , and Data Fusion for multi-modal data analysis. She develops constrained and coupled factorization methods to extract interpretable patterns in applications spanning neuroimaging, metabolomics, and mobile network analysis, with emphasis on dynamic and longitudinal data structures. Her work integrates mechanistic models with data-driven approaches to enhance biological and system understanding. Analysis of her recent publications (2024-2025) reveals a dominant trend applying tensor and coupled matrix-tensor factorizations to biomedical data for biomarker discovery, particularly in metabolomics and neuroimaging. Key innovations include tracking evolving patterns in temporal data (tPARAFAC2), constrained fusion methods (dCMF), and integration of mechanistic models with tensor decompositions for longitudinal analysis. As Head of the Department of Data Science and Knowledge Discovery, she leads research in developing novel data mining methodologies and their real-world applications at Simula Metropolitan, with significant contributions to interpretable AI for complex systems.

Takuya OouraView profile

Assistant Professor

• Numerical Analysis
• Numerical Integration
• Algorithms
• +3 more

Takuya Ooura is an Assistant Professor at the Research Institute for Mathematical Sciences (RIMS) at Kyoto University, specializing in numerical analysis and mathematical software development. His work bridges theoretical mathematics with practical applications in scientific computing and software development. Dr. Ooura earned his educational credentials through a rigorous academic path: he graduated from Hokuriku High School in 1987; completed his undergraduate studies at Nagoya University's School of Science in 1992; earned his Master of Engineering from the Department of Applied Physics at the University of Tokyo in 1994; and completed his PhD (Engineering) from the same department in 1997. His academic journey continued with a Research Fellowship from the Japan Society for the Promotion of Science (1997) followed by a position as Research Associate at RIMS, Kyoto University (2000). Dr. Ooura's research focuses on numerical integration algorithms, particularly his groundbreaking double exponential formula for Fourier-type integrals, which has been incorporated into Mathematica's NIntegrate function. He has also developed a high-speed FFT library that's utilized in Google Chrome browser (visible in chrome://credits). His work on continuous Euler transformation for accelerating convergence of slowly decaying integrals represents significant innovation in numerical analysis. His research spans both theoretical development and practical implementation of mathematical algorithms with real-world applications. His publication record demonstrates consistent contributions to numerical analysis, with particular emphasis on quadrature methods, integral transforms, and high-precision computation. His work shows a clear progression from theoretical foundations to practical implementations, with several algorithms achieving widespread adoption in commercial and open-source software. Paper prize awarded by JSIAM (2000) for 'A continuous Euler transformation and its application to Fourier transforms of slowly decaying functions' Paper prize awarded by JSIAM (2001) for 'Improvement of the PI Calculation Algorithm and Implementation of Fast Multiple-Precision Computation' Paper prize awarded by JSIAM (2005) for 'An Improved Convergence Test for the Double Exponential Formula' Japan Society for Industrial and Applied Mathematics 4th Achievement Award (2014) for 'Pioneering and practical development of the double exponential numerical integration method' Dr. Ooura has developed several widely used mathematical software packages including the double exponential integral formula, Clenshaw-Curtis numerical integration rule, and a general-purpose FFT library. His FFT package is particularly notable for its speed and accuracy, with benchmark tests showing superior performance compared to other implementations. His software has been incorporated into major projects including Google Chrome and SETI@home, demonstrating the practical impact of his theoretical work. His future research directions include further development of numerical computation libraries and applying his methods to various computational problems.

• Computer Vision
• Machine Learning
• Data-Driven Learning
• +5 more

Alexei A. Efros is the Howard Friesen Professor in the EECS Department at the University of California, Berkeley, and a core member of the Berkeley Artificial Intelligence Research (BAIR) Lab. Previously, he spent a decade at Carnegie Mellon University’s Robotics Institute. His research focuses on data-driven computer vision, self-supervised learning, computational photography, and generative models. He has pioneered advancements in visual representation learning, including seminal work on neural radiance fields and generative adversarial networks. Education Background: Efros holds a PhD in Computer Science from MIT, though specific details of his academic journey are not explicitly provided in the text. His career includes postdoctoral research at the University of Oxford with Andrew Zisserman and collaborative work with Team WILLOW at INRIA Paris. Research Interests: Efros explores how vast uncurated visual data can be leveraged for understanding and synthesizing the visual world. Key areas include self-supervised learning, generative models, and applications in robotics and art. His lab has contributed influential techniques such as Style Transfer, GAN-based image synthesis, and neural scene representation learning. Recent work emphasizes real-time adaptation (Test-Time Training), 3D perception models, and ethical AI implications of generative systems. Publications: Over 150+ publications span topics like Generative Adversarial Networks (GANs), unsupervised learning, and visual-linguistic models. Notable works include Unpaired Image-to-Image Translation (CUT/GAU), Style Transfer , and Swapping Autoencoder . His research has significant industry impact, with techniques adopted in Adobe’s software and generative AI applications. Grants & Collaborations: Efros has secured major funding from NSF, DARPA, and industry partnerships (e.g., Adobe, NVIDIA). He co-leads projects on scalable vision models, ethical AI, and real-world perception systems. Current collaborations include work with MIT, NYU, and INRIA Paris. Labs & Teams: Leads the BAIR Vision Group at Berkeley, fostering interdisciplinary research between computer vision, graphics, and robotics. The group emphasizes Slow Science principles, prioritizing deep exploration over rapid publication.

Giovanni De MicheliView profile

Professor

• Integrated Circuit Design
• Hardware Synthesis
• Logic Optimization
• +5 more

Giovanni De Micheli is a Full Professor at École Polytechnique Fédérale de Lausanne (EPFL), holding appointments in both the School of Computer and Communication Sciences (IC) and the School of Engineering (STI). He is affiliated with the Integrated Systems Laboratory (LSI) and serves as Scientific Director of EcoCloud. His primary office is located in INF 341 building at EPFL's Lausanne campus where he conducts research and teaching activities. Professor De Micheli's research spans multiple domains in electronic design automation and integrated circuit design. His primary interests include: Integrated Circuit Design and Hardware Synthesis Logic Optimization and Boolean Methods Nanoelectronics and 2D Electronics Quantum Computing and Emerging Technologies Low-Power Electronics and Energy-Efficient Systems Electronic Design Automation for Novel Computing Paradigms His recent publications demonstrate a strong focus on advancing logic synthesis techniques for both conventional and emerging technologies. Professor De Micheli's work bridges theoretical foundations with practical applications, particularly in adapting traditional EDA methods to novel computing paradigms like quantum computing and superconducting electronics. His research group has made significant contributions to the development of efficient algorithms for circuit optimization across multiple technology domains, with particular emphasis on area, power, and delay optimization while maintaining functional correctness. Professor De Micheli has supervised numerous doctoral students throughout his career at EPFL, mentoring over 40 PhD candidates whose research spans various aspects of electronic design automation, circuit design, and emerging technologies. His students have gone on to make significant contributions in both academia and industry. As Scientific Director of EcoCloud, Professor De Micheli leads research initiatives focused on energy-efficient computing systems and sustainable cloud infrastructure. His laboratory, the Integrated Systems Laboratory, serves as a hub for interdisciplinary research connecting computer science, electrical engineering, and materials science, with particular focus on next-generation computing technologies.