May 2024 - talks - lists.cs.princeton.edu

CS Colloquium Speakers: week of March 4
by Emily C. Lawrence 03 Dec '24

03 Dec '24

CS Colloquium Speakers Speaker: Sewon Min, University of Washington Date: Monday, March 4 Time: 12:30pm EST Location: CS 105 Host: Danqi Chen Event page: https://www.cs.princeton.edu/events/26579 Register for live-stream online here: https://princeton.zoom.us/webinar/register/WN_5uVGEpXNTVe8eBNSLpxKtg Title: Rethinking Data Use in Large Language Models Abstract: Large language models (LMs) such as ChatGPT have revolutionized natural language processing and artificial intelligence more broadly. In this talk, I will discuss my research on understanding and advancing these models, centered around how they use the very large text corpora they are trained on. First, I will describe our efforts to understand how these models learn to perform new tasks after training, demonstrating that their so-called in context learning capabilities are almost entirely determined by what they learn from the training data. Next, I will introduce a new class of LMs—nonparametric LMs—that repurpose this training data as a data store from which they retrieve information for improved accuracy and updatability. I will describe my work on establishing the foundations of such models, including one of the first broadly used neural retrieval models and an approach that simplifies a traditional, two-stage pipeline into one. I will also discuss how nonparametric models open up new avenues for responsible data use, e.g., by segregating permissive and copyrighted text and using them differently. Finally, I will envision the next generation of LMs we should build, focusing on efficient scaling, improved factuality, and decentralization. Bio: Sewon Min is a Ph.D. candidate in the Paul G. Allen School of Computer Science & Engineering at the University of Washington. Her research focuses on language models (LMs): studying the science of LMs, and designing new model classes and learning methods that make LMs more performant and flexible. She also studies LMs in information-seeking, legal, and privacy contexts. She is a co-organizer of multiple tutorials and workshops, including most recently at ACL 2023 on Retrieval-based Language Models and Applications and upcoming at ICLR 2024 on Mathematical and Empirical Understanding of Foundation Models. She won a paper award at ACL 2023, received a J.P. Morgan Fellowship, and was named an EECS rising star in 2022. Speaker: Lianmin Zheng, University of California, Berkeley Date: Tuesday, March 5 Time: 12:30pm EST Location: CS 105 Host: Ravi Netravali Event page: https://www.cs.princeton.edu/events/26593 Register for live-stream online here: https://princeton.zoom.us/webinar/register/WN_ghiT2HbhSPqKBouJ-VwzFg Title: Scalable and Efficient Systems for Large Language Models Abstract: Large Language Models (LLMs) have been driving recent breakthroughs in AI. These advancements would not have been possible without the support of scalable and efficient infrastructure systems. In this talk, I will introduce several underlying systems I have designed and built to support the entire model lifecycle, from training to deployment to evaluation. First, I will present Alpa, a system for large-scale model-parallel training that automatically generates execution plans unifying data, operator, and pipeline parallelism. Next, I will discuss efficient deployment systems, covering the frontend programming interface and backend runtime optimizations for high-performance inference. Finally, I will complete the model lifecycle by presenting our model evaluation efforts, including the crowdsourced live benchmark platform, Chatbot Arena, and the automatic evaluation pipeline, LLM-as-a-Judge. These projects have collectively laid a solid foundation for large language model systems, being widely adopted by leading LLM developers and companies. I will conclude by outlining some future directions of machine learning systems, such as co-optimizing across the full stack for building AI-centric applications. Bio: Lianmin Zheng is a Ph.D. student in the EECS department at UC Berkeley, advised by Ion Stoica and Joseph E. Gonzalez. His research interests include machine learning systems, large language models, compilers, and distributed systems. He builds full-stack, scalable, and efficient systems to advance the development of AI. He co-founded LMSYS.org, where he leads impactful open-source large language model projects such as Vicuna and Chatbot Arena, which have received millions of downloads and served millions of users. He also co-organized the Big Model Tutorial at ICML 2022. He has received a Meta Ph.D. Fellowship, an IEEE Micro Best Paper Award, and an a16z open-source AI grant. Speaker: Emma Dauterman, University of California, Berkeley Date: Thursday, March 7 Time: 12:30pm EST Location: CS 105 Host: Amit Levy Event page: https://www.cs.princeton.edu/events/26585 Register for live-stream online here: https://princeton.zoom.us/webinar/register/WN_soNesOKXSFml9MPaTx_1iw Title: Secure systems from insecure components Abstract: In many computer systems today, an attacker that breaks one system component can steal data from millions of users. In this talk, I will present two systems that can withstand component compromise. I will describe (1) a single sign-on system that protects user security and privacy from a compromised single sign-on server, and (2) a secure-hardware-based backup service that protects user backups from compromised secure hardware devices. These systems provide strong security and privacy properties while taking into account practical constraints such as compatibility requirements, hardware limitations, and user expectations. Each splits user secrets across different system components, using new cryptographic tools to provide necessary functionality while protecting user data. Bio: Emma Dauterman is a Ph.D. candidate at UC Berkeley where she is advised by Raluca Ada Popa and Ion Stoica. Her research interests include computer security, systems, and applied cryptography. She has received the Microsoft Research Ada Lovelace fellowship, the NSF graduate research fellowship, and a UC Berkeley EECS excellence award.

1 22

Deniz Oktay will present his FPO "Translating Between Scientific Computing and Machine Learning with Automatic Differentiation" on Tuesday, June 18, 2024 at 1:00 PM in CS 401.
by Grad Info 31 May '24

31 May '24

Deniz Oktay will present his FPO "Translating Between Scientific Computing and Machine Learning with Automatic Differentiation" on Tuesday, June 18, 2024 at 1:00 PM in CS 401 . Location: CS 401 The members of Deniz' committee are as follows: Examiners: Ryan Adams (Adviser), Benjamin Eysenbach, Sigrid Adriaenssens Readers: Elad Hazan, Szymon Rusinkiewicz A copy of his thesis is available upon request. Please email gradinfo(a)cs.princeton.edu if you would like a copy of the thesis. Everyone is invited to attend his talk. Abstract follows below: Scientific computing and machine learning, although historically separate fields, have seen much effort in unification as of recent years, especially as machine learning techniques have shown promise in scientific problems. In this thesis, I present work in the intersection of these areas, using automatic differentiation (AD) as the common language between the two. First, I present a methodological advancement in AD: Randomized Automatic Differentiation, a technique to reduce the memory usage of AD, and show that it can provide memory improvements in both machine learning and scientific computing applications. Next, I focus on mechanical design. I first describe Varmint: A Variational Material Integrator, which is a robust simulator for the statics of large deformation elasticity, using automatic differentiation as a first class citizen. Building this simulator allows us easy interoperability between machine learning and solid mechanics problems, and has been used as in several published and in submission works. I will then describe Neuromechanical Autoencoders, where we coupled neural network controllers with mechanical metamaterials to create artificial mechanical intelligence. The neural network "encoder" consumes a representation of the task-in this case, achieving a particular deformation-and nonlinearly transforms this into a set of linear actuations which play the role of the latent encoding. These actuations then displace the boundaries of the mechanical metamaterial inducing another nonlinear transformation due to the complex learned geometry of the pores; the resulting deformation corresponds to the "decoder".

1 0

Susan Tan will present her FPO "Transpilation Utilizing Language-agnostic IR and Interactivity for Parallelization" on 05/30/2024 at 1pm in Friend Center 004.
by Gradinfo 29 May '24

29 May '24

Susan Tan will present her FPO "Transpilation Utilizing Language-agnostic IR and Interactivity for Parallelization" on 05/30/2024 at 1pm in Friend Center 004. The members of her committee are as follows: Examiners: Amit Levy, Aarti Gupta, Zachary Kincaid, David August (Adviser) Readers: Johannes Doerfert (LLNL), Michael Kruse (AMD), and David August All are welcome to attend. Abstract: Migrating codes between architectures is difficult because different execution models require different types of parallelism for optimal performance. Previous approaches, like libraries or source-level tools, generate correct and natural-looking syntax for the new parallel model with limited optimization and largely leave performance engineering to the programmer. Recent approaches, such as transpilation at the compiler intermediate representation (IR) level, can automate performance engineering, but profitability can be limited by not having facts known only to the programmer. Decompiling the optimized program could leverage the strength of existing compilers to provide programmers with a natural compiler-parallelized starting point for further parallelization or refinement. Despite this potential, existing decompilers fail to do this because they do not generate portable parallel source code compatible with arbitrary compilers of the source language. This thesis provides a method for migrating code such that the compiler and programmer work together to generate code with optimal performance. To achieve this, it introduces Tulip🌷 , a source-to-source code generation framework that operates via IR-level transpilation. Transpilation at the IR level enables Tulip to generalize the transformations applied to retarget parallelism. Furthermore, Tulip integrates the state-of-the-art automatic parallelization framework to explore additional parallelism expressible only in the target parallel programming model. It then generates natural source code through a novel decompiler, SPLENDID, in a high-level parallel programming language (OpenMP), which can be interactively optimized and tuned with programmer intervention. For 19 Polybench benchmarks, Tulip-generated OpenMPCUDA offloading programs perform 14% faster than the original CUDA sources on NVIDIA GPUs. Moreover, transpilation to the CPU leads to a 2.9x speedup over the best state-of-the-art transpiler. Tulip-generated portable parallel code is also more natural than what existing decompilers produce, resulting in a 39x higher average BLEU score. This thesis includes contributions from Yebin Chon, Ziyang Xu, Sophia Zhang, and David I. August from Princeton Liberty Research Group, Brian Homerding, Yian Su, and Simone Campanoni from Northwestern Arcana Lab, Michael Kruse (AMD), Johannes Doerfert (LLNL), William S. Moses (UIUC), and Ivan R. Ivanov (Tokyo Tech).

2 1

Shaowei Zhu will present his FPO "Monotone Termination Analysis" on June 6, 2024 at 12pm in CS 105.
by Nicki Mahler 22 May '24

22 May '24

Shaowei Zhu will present his FPO "Monotone Termination Analysis" on June 6, 2024 at 12pm in CS 105. The members of his committee are as follows: Examiners: Zak Kincaid (Adviser), Andrew Appel, and Lennart Beringer Readers: David Walker, Aarti Gupta All are welcome to attend. Abstract follows below: Many decision problems, particularly those concerning program semantics, are undecidable. One prominent example is determining whether a program terminates. To address this, modern static analyzers employ heuristics to prove termination, which could be effective for certain programs but not all. Users of these analyzers struggle to understand why the tool works or not for their source programs, due to the unpredictable behavior of heuristics. For instance, annotating a loop with an invariant might make a tool that previously worked fail or not terminate. This dissertation investigates methodologies for designing termination analyses that can prove termination of real-world programs while ensuring predictability through monotonicity. A monotone analysis maintains or improves its results with additional input information, such as more loop invariants or assumptions that limit the program's state space. Monotonicity aligns with user expectations that analysis results should not deteriorate with more information given to the analysis. To achieve a monotone program analysis, we extend the algebraic program analysis framework, which uses regular expressions to represent finite program paths and interpret labels and operators within these expressions. We introduce a framework for compositional and monotone termination analysis on whole programs by incorporating an omega operator to represent infinite paths, and we develop various monotone termination analyses through different implementations of this operator based on various techniques, including analyzing phase transitions in loops, abstracting the behavior of the loop into a linear dynamical system, or synthesizing (lexicographic) polynomials ranking functions. Experimental results demonstrate that it is possible to design monotone static termination analysis that is effective for both linear and nonlinear programs and competitive with the state-of-the-art non-monotone methods in terms of the number of benchmark tasks solved and running time.

1 0

Peter Halmos will present his General Exam "DeST-OT: Alignment of Spatiotemporal Transcriptomics Data" on Friday, May 17, 2024 at 3:00 PN in CS 301 and via zoom.
by CS Grad Department 14 May '24

14 May '24

Peter Halmos will present his General Exam "DeST-OT: Alignment of Spatiotemporal Transcriptomics Data" on Friday, May 17, 2024 at 3:00 PN in CS 301 and via zoom. Zoom link: https://princeton.zoom.us/s/5942235413 Committee Members: Ben Raphael (advisor), Ellen Zhong, Olga Troyanskaya Abstract: Spatially resolved transcriptomics (SRT) measures mRNA transcripts at thousands of locations within a tissue slice, revealing spatial variations in gene expression and distribution of cell types. In recent studies, SRT has been applied to tissue slices from multiple timepoints during the development of an organism. Alignment of this spatiotemporal transcriptomics data can provide insights into the gene expression programs governing the growth and differentiation of cells over space and time. We introduce DeST-OT (Developmental SpatioTemporal Optimal Transport), a method to align SRT slices from pairs of developmental time points using the framework of optimal transport (OT). DeST-OT uses semi-relaxed optimal transport to precisely model cellular growth, death, and differentiation processes that are not well-modeled by existing alignment methods. We demonstrate the advantage of DeST-OT on simulated slices. We further introduce two metrics to quantify the plausibility of a spatiotemporal alignment: a growth distortion metric which quantifies the discrepancy between the inferred and the true cell type growth rates, and a migration metric which quantifies the distance traveled between ancestor and descendant cells. DeST-OT outperforms existing methods on these metrics in the alignment of spatiotemporal transcriptomics data from the development of mouse kidney and axolotl brain. Moreover, DeST-OT shows better correlation with the expression of annotated proliferation and apoptosis genes. Reading List: https://docs.google.com/document/d/1vYk8ZLMaOgIiMaQpQSIBh-z0KIotPOcMa6p8Nh2… Everyone is invited to attend the talk, and those faculty wishing to remain for the oral exam following are welcome to do so.

1 0

PPH Lunch seminar: Chandra Theesfeld and Rachel Kaletsky, Friday at 12 pm at 252 Nassau Street
by Emily C. Lawrence 14 May '24

14 May '24

Please join the Princeton Precision Health Initiative for the next talks in our lunch time seminar series this Friday, May 17, at 12:00 pm, in our new space at 252 Nassau Street. Chandra Theesfeld from Genomics and PPH will present a talk titled, “Unraveling the heterogeneity of complex disease: data-driven analysis of genotype, phenotype, and environment”, and Rachel Kaletsky from Genomics and Mol Bio will present a talk titled, “Defining a biological clock for female reproductive decline”. Lunch will be provided. Please note that getting to the seminar space currently requires that you climb a set of stairs. If an accommodation is needed, please contact PPH in advance at: [ mailto:princetonPPH@princeton.edu | princetonPPH(a)princeton.edu ] The PPH Team ************************************ Princeton Precision Health [ http://pph.princeton.edu/ | pph.princeton.edu ] [ mailto:princetonpph@princeton.edu | princetonpph(a)princeton.edu ]

1 0

Ann Zhou will present her General Exam " Addressing Bufferbloat & Buffer Pressure for Internet Routers through Probing" on May 16, 2024 at 12pm in Friend 125 and zoom.
by Nicki Mahler 13 May '24

13 May '24

Ann Zhou will present her General Exam " Addressing Bufferbloat & Buffer Pressure for Internet Routers through Probing" on May 16, 2024 at 12pm in Friend 125 and zoom. Zoom link https://princeton.zoom.us/j/98050139504. The members of her committee are as follows: Maria Apostolaki(adviser), Jen Rexford, Amit Levy Link to abstract & reading list: https://docs.google.com/document/d/1ab1K7uHhuzvembKjPxjTIQX6Bug8sGDH8J7Qz9g…

1 0

Cindy Zhang will present her General Exam "Dual Graphs of Crystallographic Fourier Series" on Thursday, May 16, 2024 at 3:30 PM in CS 402.
by CS Grad Department 13 May '24

13 May '24

Cindy Zhang will present her General Exam "Dual Graphs of Crystallographic Fourier Series" on Thursday, May 16, 2024 at 3:30 PM in CS 402. Committee Members: Ryan Adams (advisor), Bernard Chazelle, Benjamin Eysenbach Abstract: Many physical theories are grounded in mathematical symmetries, making it important to incorporate these symmetries into machine learning models of physical systems in order to obtain accurate and physically meaningful outputs. In materials science, the symmetries of interest are described by crystallographic groups. In physical systems where the particles satisfy the symmetries of a crystallographic group, many of the systems’ properties are described by functions invariant to these symmetries. In this work, we present a dual graph construction for the Fourier representation of functions invariant to crystallographic group symmetries. Using these graphs, we derive for each crystallographic group: (i) the Fourier bases for group-invariant functions, (ii) the irreducible representations corresponding to energy levels of a quantum-mechanical system, and (iii) the eigenvalue multiplicities of the Laplacian as a function of the sum-of-squares. The Fourier bases provide a method of embedding exact group invariances into machine learning models. Reading List: https://docs.google.com/document/d/1TyRMtOigYf2aqO5cvcV7n00l5BLIm92xH5p8xKG… Everyone is invited to attend the talk, and those faculty wishing to remain for the oral exam following are welcome to do so.

1 0

Clover Zheng will present her General Exam "Identifying multiple spatial patterns and spatially variable metabolites in spatial metabolomics data with deep learning" on Thursday, May 16, 2024 at 2:00 PM in Friend 112 and via zoom.
by CS Grad Department 13 May '24

13 May '24

Clover Zheng will present her General Exam "Identifying multiple spatial patterns and spatially variable metabolites in spatial metabolomics data with deep learning" on Thursday, May 16, 2024 at 2:00 PM in Friend 112 and via zoom. Zoom link: https://princeton.zoom.us/j/96407863507 Committee Members: Ben Raphael (advisor), Ellen Zhong, Yuri Pritykin Abstract: https://docs.google.com/document/d/1A-wTSLIdhLh_A3wRFtPl7EvxCh4-ld26vGLVcU8… Reading List: https://docs.google.com/document/d/1A-wTSLIdhLh_A3wRFtPl7EvxCh4-ld26vGLVcU8… Everyone is invited to attend the talk, and those faculty wishing to remain for the oral exam following are welcome to do so.

1 0

Jingyi Liu will present her General Exam "Profitable Deviation in Blockchain Protocol Using Cryptographic Self-Selection is Detectable" on May 15, 2024 at 3pm in Friend 005 and Zoom
by Gradinfo 13 May '24

13 May '24

Jingyi Liu will present her General Exam "Profitable Deviation in Blockchain Protocol Using Cryptographic Self-Selection is Detectable" on May 15, 2024 at 3pm in Friend 005 and Zoom Zoom: [ https://princeton.zoom.us/j/4597498062 | https://princeton.zoom.us/j/4597498062 ] Committee members: Mark Braverman (adviser), Matt Weinberg, and Pravesh Kothari Reading list: [ https://docs.google.com/document/d/1o6W4XiVHxwsTIWgP3RljxlFLhM7ChB2aFcbBASu… | https://docs.google.com/document/d/1o6W4XiVHxwsTIWgP3RljxlFLhM7ChB2aFcbBASu… ] Abstract: [ https://docs.google.com/document/d/1Nl1oyQGa4e22wh2QDiGsz1LOzHRQIW9N18BFTrG… | https://docs.google.com/document/d/1Nl1oyQGa4e22wh2QDiGsz1LOzHRQIW9N18BFTrG… ]

1 0