Sicun Gao is an Associate Professor in the Computer Science and Engineering department at the University of California, San Diego. His research focuses on practical algorithms for NP-hard search and optimization problems in computational systems, emphasizing combinatorial perspectives in numerical and statistical contexts to achieve reliable autonomy. Research Interests: Automated reasoning, Hamilton-Jacobi reachability, safe reinforcement learning, control barrier functions, and optimization in cyber-physical systems. Teaching: Courses on AI search, optimization, and graduate research seminars. The 15 most recent publications highlight advancements in safe AI control, motion planning, and policy optimization, often integrating neural networks with formal verification. Awards include the IEEE Power & Energy Society Technical Committee Prize Paper Award and the IROS RoboCup Best Paper Award. He advises PhD students working on AI-driven control and robotics, with alumni placed at institutions like Seoul National University, Amazon, and Apple. Grants include NSF Career, Air Force Young Investigator, and DARPA Assured Autonomy funding. His lab develops tools like dReal for automated reasoning in nonlinear theories over the reals.
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.
Andreas Vlachos is a Professor of Natural Language Processing and Machine Learning at the Department of Computer Science and Technology, University of Cambridge, and holds the Dinesh Dhamija Fellowship at Fitzwilliam College. His research spans dialogue modeling, automated fact-checking, imitation learning, semantic parsing, biomedical text mining, and trustworthiness in AI systems. PhD in Computer Science, University of Cambridge (supervised by Ted Briscoe and Zoubin Ghahramani) Lecturer at University of Sheffield Postdoctoral roles at UCL, University of Cambridge (NLIP group, Stephen Clark), and University of Wisconsin-Madison (Mark Craven) Current research focuses on evaluating and mitigating biases in language models, advancing fact-checking methodologies, and improving model robustness through interpolation, reinforcement learning, and causal reasoning. His work integrates natural logic, knowledge graphs, and multimodal evidence for verification tasks. Recent publications address uncertainty quantification, temporal planning benchmarks, and ethical framing of NLP artifacts. Grants from ERC, EPSRC, Facebook, Google, and the Alan Turing Institute fund his research team. Collaborations include Sebastian Riedel, Stephen Clark, and Mark Craven. Key projects explore disinformation detection, long-form generation, and confidence calibration in AI systems.
Andre Wibisono serves as Assistant Professor in Yale University's Department of Computer Science with a secondary appointment in Statistics & Data Science, joining the faculty in 2021 after postdoctoral research at University of Wisconsin-Madison and Georgia Institute of Technology. His educational background includes: Ph.D. in Computer Science, UC Berkeley M.A. in Statistics, UC Berkeley M.Eng. in Computer Science, MIT S.B. in Mathematics and Computer Science, MIT Wibisono's research focuses on algorithm design for machine learning through optimization, sampling, and game theory , leveraging dynamical systems and information theory to develop accelerated discrete-time algorithms from continuous dynamics. His work provides theoretical foundations for efficient machine learning systems with applications in generative modeling and constrained optimization. Recent publications (2023-2025) demonstrate consistent innovation in Hamiltonian-based optimization , constrained-space sampling , and min-max game convergence , characterized by rigorous mathematical analysis connecting continuous dynamics to discrete algorithms. Key trends include randomized integration for acceleration, phi-divergence convergence guarantees, and symplectic geometry applications to mirror descent. Scientific recognition includes: NSF CAREER Award for developing algorithmic frameworks bridging continuous and discrete dynamics He actively mentors current students (Siddharth Mitra, Kaylee Yang, Jane Lee, Qiang Fu, Peter Wang) and has guided two postdocs to faculty positions. Research is funded through the NSF CAREER award and collaborative CIF grants focused on Hamiltonian dynamics for sampling and optimization. His Yale research group develops theoretical foundations for next-generation machine learning algorithms, emphasizing mathematical rigor in optimization and sampling with applications to generative modeling and constrained inference problems.
Song Mei is an Assistant Professor in the Department of Statistics and Department of Electrical Engineering and Computer Sciences at the University of California, Berkeley. She received her Ph.D. from Stanford University in 2020 under Andrea Montanari and maintains active research collaborations with institutions including Amazon (as a 2023 Amazon Research Award recipient) and OpenAI (where she is currently on leave). Her research spans the intersection of statistics, machine learning, information theory, and computer science, with particular emphasis on foundational theories for modern AI systems. Key interests include language models, diffusion models, quantum algorithms, and high-dimensional statistics, often leveraging insights from statistical physics literature. Analysis of her recent publications reveals a strong focus on theoretical underpinnings of generative AI, with significant contributions to understanding contrastive pre-training (CLIP), attention mechanisms in LLMs, and mathematical foundations of diffusion models. Her work demonstrates consistent interdisciplinary connections between statistical theory and practical AI development. Sloan Research Fellowship (2025) Noether Early Career Scholar Award (2025) Google Research Scholar Award (2024) Amazon Research Award (2024) She actively advises graduate students through MA programs and has secured significant research grants including Amazon Research Awards. Her work with AGI Labs at Amazon demonstrates applied impact of theoretical research. Current projects focus on mechanistic interpretability of large language models and mathematical frameworks for generative AI. Professor Mei leads research on the statistical principles behind frontier AI models, with particular focus on developing rigorous theoretical frameworks for understanding emergent behaviors in large-scale systems.
Gabriele Farina is an Assistant Professor at MIT in the Department of Electrical Engineering and Computer Science (EECS) and the Laboratory for Information and Decision Systems (LIDS), with additional affiliations at the Operations Research Center (ORC). Holding the X-Window Consortium Career Development Chair, his research focuses on theoretical and algorithmic foundations for learning and computational decision-making under imperfect information, integrating game theory, machine learning, optimization, and statistics. He previously served as a Research Scientist at Meta's Fundamental AI Research (FAIR) group, where he contributed to Cicero, a human-level AI agent combining strategic reasoning and natural language. Ph.D. in Computer Science from Carnegie Mellon University (advisor: Tuomas Sandholm) Facebook Fellowship (2019-2020) in Economics and Computation Recipient of multiple awards including ACM SIGecom dissertation award, NSF CAREER, and AI2050 Early Career Fellow His research spans four key areas: (1) No-Regret Learning Dynamics in extensive-form games; (2) Correlation and Mediated Equilibria in sequential decision-making; (3) Team Games and Team Max-Min Equilibria; and (4) Human Modeling and Equilibrium Perfection. His work addresses challenges in scalable equilibrium computation, stability of learning algorithms, and robustness to mistakes in multi-agent systems. Recent publications highlight advancements in polynomial-time equilibrium computation, cautious optimism algorithms, and connections between regret minimization and mirror descent. These contributions appear in top venues like COLT, NeurIPS, ICML, and AAAI, with keywords spanning game theory, optimization, and machine learning. NSF CAREER award AI2050 Early Career Fellow Facebook Fellowship ACM SIGecom dissertation award GameSec 2024 best paper award ICLR 2023 outstanding paper honorable mention His research group at MIT collaborates on projects involving strategic reasoning, human-level AI agents, and equilibrium refinements, with applications to games like Diplomacy and poker. Current efforts include developing faster algorithms for correlated equilibria and exploring connections between machine learning and economic theory.
Prof. Massimo Fornasier holds the Chair of Applied Numerical Analysis at the Technical University of Munich (TUM), within the School of Computation, Information and Technology and the Department of Mathematics. His research focuses on mathematical modeling, numerical analysis, and data-driven methods, particularly in areas like compression, sparse recovery, and optimization. He has made significant contributions to consensus-based optimization, control of multiagent systems, and applications in image/signal processing. Education: PhD in Computational Mathematics, University of Padua (2003) Postdoctoral fellowships at University of Vienna, Sapienza University of Rome, and Princeton University Awards: ERC Starting Grant (2012) START Prize (2011) Prix de Boelpaepe (2009) His work bridges theoretical analysis and computational methods, with applications ranging from compressive sensing to machine learning. Recent research emphasizes consensus-based optimization frameworks and their global convergence properties. Editorial roles include journals like Networks and Heterogeneous Media and Calcolo . He leads research groups in areas such as Data Science and Numerical Analysis at TUM.
Mohammad Modarres is the Nicole J. Kim Eminent Professor at the University of Maryland within the A.J. Clark School of Engineering. He serves as Director of the Center for Risk and Reliability (CRR) and is a Professor of Nuclear Engineering in the Department of Mechanical Engineering. Dr. Modarres co-founded the world's first degree-granting graduate curriculum in reliability engineering at the University of Maryland and has established himself as an international expert in reliability and risk analysis. Dr. Modarres received his educational credentials from prestigious institutions: B.S. in Mechanical Engineering from Tehran Polytechnic M.S. in Mechanical Engineering from MIT M.S. and Ph.D. in Nuclear Engineering from MIT Dr. Modarres' research spans multiple critical areas in engineering risk and reliability. His primary interests include probabilistic risk assessment, uncertainty analysis, probabilistic physics of failure, and probabilistic fracture mechanics. His work encompasses both experimental investigations and sophisticated probabilistic model development. He has made significant contributions to materials degradation science, prognosis and health management systems, and nuclear safety analysis. His research bridges theoretical developments with practical applications in complex engineering systems, particularly in nuclear power and aerospace sectors. Analysis of Dr. Modarres' recent publications reveals a strong trend toward integrating advanced data science techniques with traditional reliability engineering. His work increasingly incorporates machine learning, deep learning, and entropy-based approaches to solve complex problems in prognostics and health management. There's a clear focus on multi-unit systems, particularly in nuclear power applications, and a growing emphasis on data-driven methodologies for remaining useful life estimation and failure prediction. His research maintains a strong foundation in probabilistic methods while embracing cutting-edge computational approaches. Dr. Modarres has received numerous prestigious honors and awards throughout his distinguished career: Nicole Y. Kim Eminent Professorship in Engineering Minta Martin Professorship in A.J. Clark School of Engineering University of Maryland Distinguished Scholar-Teacher (2019) Tommy Thompson Award for outstanding lifetime contributions to nuclear safety (American Nuclear Society) Fellow, American Nuclear Society Fellow, Institute of Electrical and Electronics Engineers (IEEE) Life Fellow of IEEE 1996 Maryland Inventor of the Year Award (in Information Sciences) FDA Commissioner Special Citation for Contributions to Risk Assessment Methods (2004) 2008 International Research Leadership Award (Society for Reliability Engineering, Quality and Operations Management) Honorary Doctorate from Universidad Da Vinci de Guatemala As the founding director of the Center for Risk and Reliability, Dr. Modarres has built a world-renowned program that has awarded over 500 Ph.D. and master's degrees. His research has been supported by significant grants from government agencies and industry partners, particularly in nuclear safety, aerospace reliability, and critical infrastructure protection. He has mentored numerous students who have gone on to become leaders in reliability engineering across various industries. His center collaborates extensively with the International Atomic Energy Agency (IAEA) and other international organizations on risk assessment methodologies. The Center for Risk and Reliability (CRR), which Dr. Modarres directs, serves as a hub for multidisciplinary research in risk and reliability engineering. The center has recently renovated its facilities to enhance collaboration among researchers from different engineering disciplines. CRR maintains strong partnerships with industry leaders including Amazon Lab126 (as evidenced by their collaboration on device durability research) and has been instrumental in advancing probabilistic risk assessment tools used in nuclear power plant safety analysis. The center hosts regular seminars, workshops, and international conferences, positioning itself at the forefront of risk and reliability research globally.
Francis Bach is a Professor and researcher at INRIA, leading the SIERRA project-team since 2011, which is part of the Computer Science Department at Ecole Normale Supérieure (ENS) within PSL Research University. His work bridges CNRS, ENS, and INRIA as a joint research effort. Elected to the French Academy of Sciences in 2020, he currently runs the ERC project SEQUOIA following his previous ERC project SIERRA (2009-2014). His research spans statistical machine learning with focus on optimization, sparse methods, kernel-based learning, neural networks, graphical models, and signal processing. Bach completed his Ph.D. in Computer Science at U.C. Berkeley under Professor Michael Jordan, followed by work at Ecole des Mines de Paris and the WILLOW project-team at INRIA/ENS/CNRS (2007-2010). His recent book "Learning Theory from First Principles" was published by MIT Press in December 2024. Bach's publication record shows consistent high-impact contributions across machine learning theory and applications, with recent work focusing on conformal prediction, diffusion models, optimization theory, and learning theory foundations. His research demonstrates strong connections between theoretical guarantees and practical algorithms, with applications spanning generative modeling, robust optimization, and statistical inference. Elected to French Academy of Sciences (2020) ERC project SIERRA (2009-2014) ERC project SEQUOIA (current) Author of "Learning Theory from First Principles" (MIT Press, 2024) Bach actively mentors numerous PhD students and postdocs, with many alumni now holding faculty positions at institutions like EPFL, Ecole Polytechnique, University of Washington, and University of Montreal. His teaching includes advanced courses on learning theory at ENS's Master's programs. He regularly presents tutorials at major conferences including COLT, NeurIPS, and ICML, demonstrating his leadership in the theoretical machine learning community.
Bryon Aragam is an Associate Professor of Econometrics and Statistics and Robert H. Topel Faculty Scholar at the University of Chicago Booth School of Business. His research focuses on the intersection of causality, statistical machine learning, and probabilistic modeling, with particular emphasis on applications to artificial intelligence systems including large language models like ChatGPT and generative models like DALL-E. Dr. Aragam completed his PhD in Statistics and a Masters in Applied Mathematics at UCLA, where he was an NSF graduate research fellow. Prior to joining the University of Chicago, he was a project scientist and postdoctoral researcher in the Machine Learning Department at Carnegie Mellon University. Research Focus: Causal structure learning in probabilistic generative models Key Areas: Causal machine learning, deep generative models, latent variable models, statistical learning theory Applications: AI interpretability, ethics, and fairness in artificial intelligence systems Teaching: Business Statistics, Econometrics and Statistics Colloquium His recent publications demonstrate a strong theoretical foundation combined with practical applications, particularly in understanding and improving AI systems. His work spans causal discovery, graphical models, deep learning, and latent variable modeling, with particular attention to the theoretical properties of these methods and their applications to real-world AI challenges. The research shows a progression toward increasingly complex problems in causal representation learning and AI interpretability. Scientific Awards: Robert H. Topel Faculty Scholar NSF Graduate Research Fellow Dr. Aragam's work has been published in top statistics and machine learning venues including the Annals of Statistics, Neural Information Processing Systems (NeurIPS), the International Conference on Machine Learning (ICML), and the Journal of Machine Learning Research (JMLR). His research group publishes broadly across both statistical and machine learning communities, demonstrating the interdisciplinary nature of his work at the intersection of statistics, machine learning, and causal inference. As a data science consultant for technology and marketing firms, Dr. Aragam has applied his expertise to problems in survey design, customer retention, logistics, and ranking, bridging the gap between theoretical research and practical applications.
Yuxin Chen is a Professor at the University of Pennsylvania , holding joint appointments in the Department of Statistics and Data Science and the Department of Electrical and Systems Engineering . Prior to UPenn, he was an Assistant Professor at Princeton University (2017-2021) and a Postdoctoral Researcher at Stanford University (2015-2017). His research spans statistics, optimization, reinforcement learning theory, diffusion models, and information theory , with a focus on theoretical foundations and practical algorithms for machine learning. Education : Ph.D. in Electrical Engineering (Stanford, 2015), M.S. in Statistics (Stanford, 2013), M.S. in Electrical and Computer Engineering (UT Austin, 2010), B.E. in Electrical/Microelectronics (Tsinghua, 2008). Research Interests encompass theoretical and applied aspects of machine learning, including nonconvex optimization , sample complexity analysis , low-dimensional adaptation , and generative modeling . His work bridges mathematical rigor with real-world applications, particularly in scientific imaging and high-dimensional data analysis. Scientific Awards include the SIAM Activity Group on Imaging Science Best Paper Prize (2024) Alfred P. Sloan Fellowship (2022) NSF Career Award (2022) Google Research Scholar Award (2022) IEEE Transactions on Power Electronics Prize Paper Award (2024) Advising and Grants : He has mentored numerous students who have transitioned to academic roles at institutions like UIUC and UW-Madison. His research is supported by grants from the NSF , Amazon , and Google , with recent projects focusing on controllable diffusion models and efficient reinforcement learning algorithms .
Amir-massoud Farahmand is an Associate Professor at the Polytechnique Montréal (Department of Computer and Software Engineering) and a Status-Only Associate Professor at the University of Toronto (Department of Computer Science). He is also a Core Academic Member at Mila (Quebec AI Institute). His research focuses on computational and statistical mechanisms for designing efficient reinforcement learning (RL) agents and adaptive algorithms. Dr. Farahmand's research spans reinforcement learning, optimal transport, adversarial robustness, and model-based methods. He has extensively studied regularization in RL, distributional approaches, and algorithm design for stability and convergence. His textbook Lecture Notes on Reinforcement Learning (2021) emphasizes mathematical intuition over algorithmic collections. Recent publications highlight trends in high-update-ratio RL, distributional equivalence, and self-prediction for task understanding. He is actively involved in teaching, having previously instructed courses on machine learning, neural networks, and RL at the University of Toronto. Scientific Awards : Ontario Early Researcher Award (2024) for Accelerated Reinforcement Learning Algorithms Dr. Farahmand has mentored numerous students, including his first PhD graduate Yangchen Pan (now at Oxford) and MSc students like Allen Bao (AMD) and Farnam Mansouri (University of Waterloo). He is currently recruiting graduate students at Polytechnique Montréal and Mila for 2025 admissions.
Laurens Lootens is a Researcher in the Department of Applied Mathematics and Theoretical Physics (DAMTP) at the University of Cambridge. His work focuses on theoretical physics, particularly in quantum lattice models, topological phases of matter, and mathematical structures underlying quantum systems. He is affiliated with the High Energy Physics research group within DAMTP. His research interests include dualities in quantum systems, matrix product operator symmetries, conformal field theories, and tensor network methods. Lootens explores topics such as entanglement in many-body systems, symmetry-protected topological phases, and the interplay between algebraic structures and physical phenomena. Publications highlight his contributions to understanding lattice representations of dualities, topological sectors in quantum models, and critical lattice models for conformal field theories. His work bridges theoretical frameworks with computational methods, advancing both fundamental physics and quantum information science.
Karl-Theodor Sturm is a Professor of Mathematics at the University of Bonn, holding this position since 1997. He is affiliated with the Institute for Applied Mathematics and leads the Cluster of Excellence Hausdorff Center for Mathematics. His academic journey includes a PhD (1989) and habilitation (1993) from the University of Erlangen-Nürnberg, followed by postdoctoral positions at Zurich, Erlangen-Nürnberg, and the Max Planck Institute for Mathematics in the Sciences (MPI Leipzig). He has held visiting professorships at Stanford, Toulouse, Paris, and Bonn. Sturm's research focuses on stochastic analysis and geometric analysis, particularly in optimal transport, metric measure spaces, synthetic curvature bounds, and diffusion processes. His work on synthetic Ricci curvature bounds, developed in competition with Cédric Villani, has been highly influential. He received the ERC Advanced Grant (2016-2022) for research on metric measure spaces and Ricci curvature, and was a Plenary Speaker at the 2020 European Congress of Mathematics. His leadership roles include Vice Chairman of Collaborative Research Center SFB 611 (2002–2012), Managing Director of the Institute for Applied Mathematics (2007–2010), and Coordinator of the Hausdorff Center for Mathematics (2012–2019). Awards include the Heisenberg Fellowship (1994) and recognition through numerous invited lectures and editorial roles. His mentorship has shaped the careers of prominent researchers such as Nicola Gigli and Jan Maas.
Piotr Zwiernik is an Associate Professor in the Department of Statistical Sciences at the University of Toronto's Faculty of Arts and Science, with a cross-appointment in the Department of Mathematics. Currently on leave from the University of Toronto, he is based in Barcelona following his return in July 2025. His academic journey includes a PhD in Statistics from the University of Warwick (2011), research positions at prestigious institutions including Mittag Leffler Institute, IPAM, TU Eindhoven, UC Berkeley, and the University of Genoa, and an Assistant Professorship at Universitat Pompeu Fabra in Barcelona (2016-2021). His research spans the intersection of statistics, mathematics, and computational methods, with particular emphasis on graphical models, covariance matrix estimation, convex analysis, tensors, and algebraic and combinatorial methods in statistics. Zwiernik's work demonstrates a consistent focus on high-dimensional statistics, mathematical statistics, and elegant theoretical frameworks that bridge abstract mathematics with practical statistical applications. His recent publications reveal a deepening exploration of tensor analysis, algebraic statistics, and the geometric properties of statistical models. Zwiernik serves as an associate editor for leading journals including Biometrika, Scandinavian Journal of Statistics, and Algebraic Statistics. His research program includes the development of the GOLAZO R package for asymmetric regularization of log-likelihood in Gaussian graphical models. As an academic leader, he has served as Associate Chair for Research in his department and actively participates in numerous international conferences and workshops, reflecting his significant standing in the statistical community. His recent publications show a strong trend toward algebraic and geometric approaches to statistical problems, with increasing focus on tensor methods, positivity constraints in statistical models, and the theoretical foundations of graphical models. The work demonstrates remarkable continuity in exploring the mathematical structures underlying statistical models while adapting to emerging challenges in high-dimensional data analysis. Zwiernik is committed to mathematical accessibility and education, guided by Federico Ardila's four axioms which emphasize equitable distribution of mathematical potential, joyful mathematical experiences, mathematics as a malleable tool, and treating every student with dignity and respect. He actively seeks PhD students with strong mathematical backgrounds for research at UPF or the Institute of Mathematics of UPC.
