CEMSE Weekly Updates - October 6, 2026 Tue, Oct 6 2026 Newsletter Upcoming Events Stay informed about the upcoming events and the latest news from CEMSE. Transformers from Compressed Representations Juan Carlos L. Alcazar, Research Scientist, Electrical and Computer Engineering Oct 11, 12:00 - 13:00 B9 R2325 transformers representation learning TEMPEST visual computing AI machine learning This talk explores how transformers can effectively be trained on compressed byte streams in order to gain computational and memory efficiencies. From Abstraction to Acceleration: Productive, Precision-Aware Scientific Computing Rabab Alomairy, Ibn Rushd Assistant Professor, Computer Science Oct 12, 12:00 - 13:00 B9 R2325 programming abstractions algorithms software/hardware co-design scientific computing HPC This talk explores how high-level programming abstractions, precision-aware algorithms, and hardware-conscious design can work together to accelerate scientific computing while improving productivity, scalability, and energy efficiency. Unbiased and Multilevel Monte Carlo Methods for Parameter Inference in Latent Stochastic Systems Miguel Angel Alvarez Ballesteros, Ph.D. Student, Applied Mathematics and Computational Science Oct 14, 11:00 - 13:00 B2 R5209 Monte carlo methods Monte Carlo stochastic optimization markov chains approximation This thesis develops efficient Monte Carlo methods for score-based parameter inference in such systems, mainly hidden Markov models driven by diffusion processes, and for the related problem of conditional stochastic optimization. Solvers That Don't Choose Sides Esmail Abdul Fattah, Postdoctoral Research Fellow, Statistics Oct 15, 12:00 - 13:00 B9 R2325 sparse computation scientific computing programming abstractions dense linear algebra Sparse Linear Algebra sTiles In this talk, I will present sTiles, a direct solver that decides sparse or dense tile by tile instead of once for the whole matrix, covering the full range from sparse to fully dense and outpacing established sparse direct solvers on the repeated factorizations that drive large-scale inference.
