- talks - lists.cs.princeton.edu

NextG Distinguished Seminar Series: Arjuna Madanayake, NEW DATE Oct 31
by Emily C. Lawrence 24 Oct '24

24 Oct '24

NextG Distinguished Seminar Series Speaker: Arjuna Madanayake, Florida International University Day: October 31, 2024 Time: 10 :30 a m Location: B205 Engineering Quadrangle Hosts: Yasaman Ghasempour and Kaushik Sengupta Event Page: [ https://ece.princeton.edu/events/spectre-signal-processing-efficient-comput… | https://ece.princeton.edu/events/spectre-signal-processing-efficient-comput… ] Title: SPECTRE: Signal Processing with Efficient Compute for Transmit and Receive Environments Abstract: Wideband radio frequency (RF) array processors form an integral building block of modern wireless communications, location, sensing, and electronic warfare systems. There is interest in multifunctional and frequency agile antenna arrays with fully digital beam processing and machine learning (ML) algorithms at the backend that operate across a wide span of frequencies starting from VHF/UHF up to mm-waves and even sub-THz ranges. Digital signal processing (DSP) across massive bandwidths – today up to about 30 GHz for some sub-THz applications – requires careful consideration of multiple design constraints. Most array DSP operations consist of typical functions, such as discrete Fourier transforms (DFTs), finite impulse response (FIR) filters, cross-correlators, multiply-accumulators, and linear algebra (LA) functions like matrix inversion. Arithmetic complexity of basic DSP algorithms must be carefully factored into their design. A classic example is the use of sparse matrix factorizations on the DFT matrix leading to the suite of fast algorithms known as Fast Fourier Transforms (FFTs) of which the most famous is the Cooley-Tukey FFT invented at Princeton and IBM. The claim to fame for the FFT is its enormously useful factorizations having low arithmetic complexity at O(N log2 N), which is tiny compared to the case of the DFT. The DFT is computed brute-force at arithmetic complexity O(N2) where N is the number of points used. The theoretical lower bounds of DFT complexity have been rigorously proven decades in the past by Heideman and Burrus. Luckily, we can get around these bounds via a sequence of approximations. We achieve O(N) complexity for FFT-like operations by recognizing how real-world RF systems have multiple sources of error. The theoretical complexity bounds of the DFT could be circumvented by computing judiciously tailored low-complexity DFT matrix approximations capable of closely preserving mathematical properties of the exact DFT. In this talk, we discuss the use of approximate DFTs (ADFT) with corresponding sparse factorizations that are free of multiplications. The author and his collaborator RJ Cintra have developed an approach for superfast wideband RF signal processing for FFT applications, including multi-beam spacetime orthogonal beamforming, orthogonal frequency division multiplexing (OFDM) and maximally-decimated uniform-DFT polyphase FIR filterbanks, that replace O(N log2 N) FFTs of Cooley-Tukey fame with factorized ADFTs that are within 2 dB of performance of exact DFT but show much lower arithmetic complexity at just O(N). This extra log2 N reduction beyond what is possible from FFTs – which of course comes at a small price – has massive benefits in both chip area and power consumption for many wideband RF systems. In this talk, the author discusses our recent DARPA, ONR, and NSF funded research activities surrounding low-complexity low-power DSP for real-world RF multi-beam beamforming, low-SWaP spatio-temporal multidimensional spectrum sensing, and low-complexity adaptive digital null-forming using Howells-Applebaum adaptive arrays that make use of the O(N) ADFTs. The talk covers the basic mathematical theory, algorithms, RF/microwave systems, digital compute and FPGA/RF-SoC and chiplet based CMOS realizations underway at Florida International University. Bio: Arjuna Madanayake is an Associate Professor of Electrical and Computer Engineering at Florida International University (FIU), Miami, Florida. He runs the RAND Lab which has 15 PhD students at present. His area of specialization is multidimensional systems and signal processing. Dr. Madanayake’s wider research interests span one- and multi-dimensional signal processing and filter design (both analog and digital), RF/microwave systems, analog-CMOS and analog computer architecture, digital systems and FPGA/RF-SoC systems, fast algorithms and digital ASIC, digital hardware based ``software defined’’ radios, electromagnetics and sub-THz nearfield processing, wireless communications, 5G/6G and mm-wave systems, spectrum, RF artificial intelligence and machine learning (AI/ML) and semiconductor technologies.

1 0

PPH Seminar: “AI in Medical Care: Deus or Machina?” TODAY at 12pm
by Emily C. Lawrence 18 Oct '24

18 Oct '24

Please join the Princeton Precision Health (PPH) Initiative for the next talk of our fall semester series on Friday, October 18, at 12:00 pm, at 252 Nassau Street. William Hanson, MD, Chief Medical Information Officer and Vice President, Hospital of the University of Pennsylvania, will present a talk, “AI in Medical Care: Deus or Machina?” Abstract: In the last 20 years, spurred by governmental incentives (American Recovery and Reinvestment) the digitization of devices and widespread adoption of electronic record keeping systems has transformed medical care for better (and some would argue for worse). Medical care and research has spawned the development of vast troves of digitized clinical and research data. Medical information flows through countless visible and invisible conduits, and has become the new oil, coveted by large companies, research initiatives and governments. This talk will describe the digitization of medicine, the players, the intended and unintended consequences and the potential benefits of AI and precision medicine. It will offer a perspective on what has gone right, what wrong and what entities like Princeton Precision Health can bring to the table. 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 ]

1 0

Matthew Weaver will present his FPO "Bicubical Directed Type Theory" on October 31, 2024 at 11am in CS 301.
by Gradinfo 17 Oct '24

17 Oct '24

Matthew Weaver will present his FPO "Bicubical Directed Type Theory" on October 31, 2024 at 11am in CS 301. The members of his committee are as follows: Andrew Appel (Advisor) Daniel Licata (Advisor/Wesleyan University) Zachary Kincaid (Reader) David Walker (Examiner) Aarti Gupta (Examiner) Title: Bicubical Directed Type Theory Directed type theory is an analogue of homotopy type theory where types represent categories, generalizing groupoids. A classical, bisimplicial approach to directed type theory, developed by Riehl and Shulman, is based on equipping each type with both a notion of path and a separate notion of directed morphism. In this setting, a directed analogue of the univalence axiom asserts that there is a universe of covariant discrete fibrations whose directed morphisms correspond to functions—a higher-categorical analogue of the category of sets and functions. While this approach suffices for formalizing mathematics, for applications to computer science we would like a computational interpretation of directed type theory. In this thesis we give a constructive model of a directed type theory with directed univalence in bicubical sets. We formalize much of this in Agda following the approach of Orton and Pitts. First, building on the cubical techniques used to give computational models of homotopy type theory, we show that there is a universe of covariant discrete fibrations with a partial directed univalence principle asserting that functions are a retract of morphisms in this universe. To complete this retraction into an equivalence, we refine the model using Coquand, Ruch, and Sattler’s work on constructive sheaf models. We introduce the cobar modality and by restricting to fibrant types that are also cobar modal, we are able to complete our construction of directed univalence. We then describe a generalization of the fibrant Riehl-Shulman extension types. We prove this in the setting of an arbitrary presheaf category with respect to a new notion of fibrancy that is given by a generic filling problem. This abstraction is general enough to capture all of the current presheaf models of type theories and their classifications of types specified by filling problems. In addition, this result extends the potential syntax of these type theories to be able to internally express any of these filling problems as fibrant types. We use this to then define a type theory in which the user can internally define classifying universes for any such filling problem. Lastly, we overview our implementation of bicubical directed type theory.

1 0

NextG Webinar: Maria Apostolaki, Oct 22 at 12pm
by Emily C. Lawrence 15 Oct '24

15 Oct '24

Faculty Spotlight Series Featured Faculty: Professor Maria Apostolaki Date: October 22, 2024 Time: 12:00 pm Speaker; Maria Apostolaki, Princeton University Registration: Participate remotely by visiting [ https://princeton.zoom.us/meeting/register/tJYucOGqrD4oGdU6LE3hruA8tNc_RD3J… | this link ] Maria Apostolaki is an Assistant Professor of Electrical and Computer Engineering at Princeton University. Her research draws from networking, security, and machine learning. Overall, her goal is to design and build networked systems that are secure, reliable, and performant. Maria has been named a rising star in Computer Networking and Communications and has received the Google Research Scholar Award, Commendation for Outstanding Teaching, and the IETF/IRTF Applied Networking Research Prize. Maria completed her PhD in 2021 at ETH Zurich. During her studies, she was a visiting student at MIT and worked at Microsoft Research and Google. After her PhD, she spent a year at Carnegie Mellon University as a postdoctoral researcher. Prior to her PhD, she earned her diploma in Electrical and Computer Engineering at the National Technical University of Athens, Greece. Professor Apostolaki will present her research on: “Combining ML and Formal Methods for network management.” The presentation will be followed by a Q&A/discussion period open to the audience. The NextG Faculty Spotlight Series is hosted by Princeton NextG, an initiative of Princeton’s School of Engineering and Applied Sciences. The NextG Initiative at Princeton is creating the foundation for intelligent networks of the future across wireless, backbone networking, and cloud systems. The initiative focuses on cross-disciplinary approaches, bringing theory to practice, and covering the ‘full stack’ from the underlying technological fabric in integrated electronic and photonic circuits and systems, edge networks, IOT and cloud, to foundational theory, algorithms and AI approaches that make these networks scalable, efficient, secure, and accessible. Zoom registration required : [ https://princeton.zoom.us/meeting/register/tJYucOGqrD4oGdU6LE3hruA8tNc_RD3J… | https://princeton.zoom.us/meeting/register/tJYucOGqrD4oGdU6LE3hruA8tNc_RD3J… ]

1 0

PPH Seminar Fri Oct 18 “AI in Medical Care: Deus or Machina?” William Hanson, MD / Registration open for Nov 7 PPH-NOAA health & climate symposium
by Emily C. Lawrence 14 Oct '24

14 Oct '24

Please join the Princeton Precision Health (PPH) Initiative for the next talk of our fall semester series on Friday, October 18, at 12:00 pm, at 252 Nassau Street. William Hanson, MD, Chief Medical Information Officer and Vice President, Hospital of the University of Pennsylvania, will present a talk, “AI in Medical Care: Deus or Machina?” Abstract: In the last 20 years, spurred by governmental incentives (American Recovery and Reinvestment) the digitization of devices and widespread adoption of electronic record keeping systems has transformed medical care for better (and some would argue for worse). Medical care and research has spawned the development of vast troves of digitized clinical and research data. Medical information flows through countless visible and invisible conduits, and has become the new oil, coveted by large companies, research initiatives and governments. This talk will describe the digitization of medicine, the players, the intended and unintended consequences and the potential benefits of AI and precision medicine. It will offer a perspective on what has gone right, what wrong and what entities like Princeton Precision Health can bring to the table. 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 ] ********************************** [ https://pph.princeton.edu/events/data-driven-precision-health-integrating-g… | Registration ] is now open for PPH-NOAA November 7 conference: "Data-Driven Precision Health: Integrating Genomics, Climate, and the Environment.” The conference, a collaboration between Princeton Precision Health and NOAA’s Geophysical Fluid Dynamics Laboratory and co-sponsored by Princeton’s High Meadows Environmental Institute , will bring together an interdisciplinary community from academia, government, and industry to explore research questions at the intersection of human health, climate, and environment. The event aims to spark discussions and inspire new research directions on future health impacts, with a focus on predicting outcomes and developing equitable risk mitigation strategies. The conference will be held in Maeder Hall, at the Andlinger Center for the Energy and the Environment. Keynote speakers include Jane Burns , Director of the Kawasaki Disease Research Center at UC San Diego and whose work on Kawasaki Disease, environment, and climate has been featured in the [ https://www.nytimes.com/2024/02/27/health/kawasaki-disease-cause.html | NY Times ] , and Juli Trtanj , Director of the NOAA Climate and Health Program and NOAA One Health Lead and co-chair of Global Heat Health Information Network (GHHIN) Management Committee . In addition, Provost Jennifer Rexford and Dean For Research Peter Schiffer will be giving remarks at the conference. Registration is free but required.

1 0

PPH Seminar: TODAY, 12pm @ 252 Nassau St
by Emily C. Lawrence 11 Oct '24

11 Oct '24

Please join the Princeton Precision Health (PPH) Initiative for the next talk of our fall semester series TODAY, October 11, at 12:00 pm, at 252 Nassau Street. Zeynep Tufekci, Henry G. Bryant Professor of Sociology and Public Affairs, will present a talk, “The Pandemic: Science, Policy and Trust.” Abstract: A key tenet of dealing with a pandemic is remembering that it is as much a communication emergency as much as a medical crisis, or so warns us the C.D.C. in its pre-pandemic publications. A pandemic communication strategy should be transparent, "acknowledge uncertainties and do not over-reassure or overpromise", and be first, right and credible, we are also reminded in various manuals. Easier said than done, of course, and especially challenging in an era of polarization, social media and charged politics. And yet, was it all inevitably doomed? There has been much memory-holing and rewriting of the pandemic history, especially erasing the uncertainty of the early days as well as the extensive cooperation and trust that was extended from the public, globally. This talk will examine the series of challenges, twists and missteps that resulted in a situation where we are, arguably, less ready to face another health crisis should one occur — such as an H5N1 outbreak, and recenter the role of trust in public policy, especially health policy. 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

Akhil Jakatdar will present his General Exam "Inferring Phylogenetic Trees from Longitudinal Single-Cell Tumor Samples" on Friday, October 11, 2024 at 3:30 PM in CS 402.
by CS Grad Department 08 Oct '24

08 Oct '24

Akhil Jakatdar will present his General Exam "Inferring Phylogenetic Trees from Longitudinal Single-Cell Tumor Samples" on Friday, October 11, 2024 at 3:30 PM in CS 402. Committee Members: Ben Raphael (advisor), Mona Singh, Yuri Pritykin Abstract: Recent advances in high-throughput single-cell sequencing technology provide the opportunity to better understand the tumor evolutionary process through large-scale longitudinal profiling studies. These studies allow for cancer progression to be observed in the same patient across multiple timepoints thus introducing a new modality of temporal information associated with the already collected genomic data. As cancer proliferates through the cell division process, we can model this evolutionary process with phylogenetic trees. We demonstrate that current single-cell tumor phylogenetic inference methods infer trees that are discordant with the additional temporal information collected through longitudinal studies. We model a longitudinal observance process underlying these single-cell longitudinal profiling studies. We integrate this longitudinal observance process model with the Perfect Phylogeny mutation evolutionary model to construct the Longitudinally Observed Perfect Phylogeny (LOPP) model which characterizes the tumor evolutionary process in longitudinal studies. As high-throughput single-cell sequencing technology is error-prone, we formulate the Maximum Likelihood Longitudinal Phylogeny Reconstruction (ML-LPR) problem that asks to reconstruct the highest likelihood phylogeny given observed single-cell sequencing reads under some probabilistic model. We prove that the ML-LPR under the LOPP model is NP-hard and introduce an ILP method, Phyllochron, that solves a variant of the ML-LPR on a fixed mutation tree. We show that Phyllochron produces longitudinal observed phylogenies on simulated data cases where established single-cell phylogeny inference methods that do not explicitly model temporal information cannot. On real data, Phyllochron infers more plausible phylogenies compared with competing methods, suggesting that the LOPP can accurately describe the tumor evolutionary process even in the event of sampling concerns and sequencing errors. Reading List: https://docs.google.com/document/d/1xOMfVhdcSmnmGLZMyhiauNsd_7KXJwIyTha79NZ… 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

NextG Distinguished Seminar Series: Thursday Oct. 10 at 4:30 pm
by Emily C. Lawrence 08 Oct '24

08 Oct '24

Speaker: Arjuna Madanayake, Florida International University Day: October 10, 2024 Time: 4:30pm Location: B205 Engineering Quadrangle Hosts: Yasaman Ghasempour and Kaushik Sengupta Event Page: [ https://ece.princeton.edu/events/spectre-signal-processing-efficient-comput… | https://ece.princeton.edu/events/spectre-signal-processing-efficient-comput… ] Title: SPECTRE: Signal Processing with Efficient Compute for Transmit and Receive Environments Abstract: Wideband radio frequency (RF) array processors form an integral building block of modern wireless communications, location, sensing, and electronic warfare systems. There is interest in multifunctional and frequency agile antenna arrays with fully digital beam processing and machine learning (ML) algorithms at the backend that operate across a wide span of frequencies starting from VHF/UHF up to mm-waves and even sub-THz ranges. Digital signal processing (DSP) across massive bandwidths – today up to about 30 GHz for some sub-THz applications – requires careful consideration of multiple design constraints. Most array DSP operations consist of typical functions, such as discrete Fourier transforms (DFTs), finite impulse response (FIR) filters, cross-correlators, multiply-accumulators, and linear algebra (LA) functions like matrix inversion. Arithmetic complexity of basic DSP algorithms must be carefully factored into their design. A classic example is the use of sparse matrix factorizations on the DFT matrix leading to the suite of fast algorithms known as Fast Fourier Transforms (FFTs) of which the most famous is the Cooley-Tukey FFT invented at Princeton and IBM. The claim to fame for the FFT is its enormously useful factorizations having low arithmetic complexity at O(N log2 N), which is tiny compared to the case of the DFT. The DFT is computed brute-force at arithmetic complexity O(N2) where N is the number of points used. The theoretical lower bounds of DFT complexity have been rigorously proven decades in the past by Heideman and Burrus. Luckily, we can get around these bounds via a sequence of approximations. We achieve O(N) complexity for FFT-like operations by recognizing how real-world RF systems have multiple sources of error. The theoretical complexity bounds of the DFT could be circumvented by computing judiciously tailored low-complexity DFT matrix approximations capable of closely preserving mathematical properties of the exact DFT. In this talk, we discuss the use of approximate DFTs (ADFT) with corresponding sparse factorizations that are free of multiplications. The author and his collaborator RJ Cintra have developed an approach for superfast wideband RF signal processing for FFT applications, including multi-beam spacetime orthogonal beamforming, orthogonal frequency division multiplexing (OFDM) and maximally-decimated uniform-DFT polyphase FIR filterbanks, that replace O(N log2 N) FFTs of Cooley-Tukey fame with factorized ADFTs that are within 2 dB of performance of exact DFT but show much lower arithmetic complexity at just O(N). This extra log2 N reduction beyond what is possible from FFTs – which of course comes at a small price – has massive benefits in both chip area and power consumption for many wideband RF systems. In this talk, the author discusses our recent DARPA, ONR, and NSF funded research activities surrounding low-complexity low-power DSP for real-world RF multi-beam beamforming, low-SWaP spatio-temporal multidimensional spectrum sensing, and low-complexity adaptive digital null-forming using Howells-Applebaum adaptive arrays that make use of the O(N) ADFTs. The talk covers the basic mathematical theory, algorithms, RF/microwave systems, digital compute and FPGA/RF-SoC and chiplet based CMOS realizations underway at Florida International University. Bio: Arjuna Madanayake is an Associate Professor of Electrical and Computer Engineering at Florida International University (FIU), Miami, Florida. He runs the RAND Lab which has 15 PhD students at present. His area of specialization is multidimensional systems and signal processing. Dr. Madanayake’s wider research interests span one- and multi-dimensional signal processing and filter design (both analog and digital), RF/microwave systems, analog-CMOS and analog computer architecture, digital systems and FPGA/RF-SoC systems, fast algorithms and digital ASIC, digital hardware based ``software defined’’ radios, electromagnetics and sub-THz nearfield processing, wireless communications, 5G/6G and mm-wave systems, spectrum, RF artificial intelligence and machine learning (AI/ML) and semiconductor technologies.

1 1

Machine Learning Seminar: Prof. Ramya Korlakai Vinayak visiting on Friday, Oct 4th
by Emily C. Lawrence 03 Oct '24

03 Oct '24

Machine Learning Seminar Prof. Ramya Korlakai Vinayak from UW-Madison is visiting this Friday, Oct 4th from 11:00 am -12:00 pm at Friend Center 008. Title: Towards Pluralistic Alignment: Foundations for Learning from Diverse Human Preferences Abstract: Large pre-trained models trained on internet-scale data are often not ready for safe deployment out-of-the-box. They are heavily fine-tuned and aligned using large quantities of human preference data, usually elicited using pairwise comparisons. While aligning an AI/ML model to human preferences or values, it is important to ask whose preference and values we are aligning it to ? The current approaches of alignment are severely limited due to their inherent uniformity assumption. While there is rich literature on learning preferences from human judgements using comparison queries, the models often focus on learning average preference over the population due to the limitations on the amount of data available per individual or on learning an individual's preference using a lot of queries. Furthermore, the knowledge of the metric, i.e., the way humans judge similarity and dissimilarity, is assumed to be known which does not hold in practice. We aim to overcome these limitations by building mathematical foundations for learning from diverse human preferences. In this talk, I will present, PAL, a personalize-able reward modelling framework for pluralistic alignment, which captures diversity in preferences while also capturing commonalities that can be learned by pooling together data from individuals. I will also discuss some recent theoretical results on per user sample complexity for generalization and fundamental limitations when there are limited pairwise comparisons. Based on work with Daiwei Chen, Yi Chen, Aniket Rege, Zhi Wang, Geelon So, Greg Canal, Blake Mason, Gokcan Tatli, and Rob Nowak. References: 1. [ https://arxiv.org/pdf/2406.08469 | PAL: Pluralistic Alignment Framework for learning from heterogeneous preferences ] (preprint, 2024) 2. [ https://arxiv.org/abs/2207.03609 | One-for-all: Simultaneous metric and preference learning ] (appeared in Neurips 2022) 3. [ https://arxiv.org/pdf/2403.19629 | Metric learning via limited pairwise comparisons ] (appeared in UAI 2024), and 4. [ https://proceedings.mlr.press/v238/tatli24a.html | Learning Populations of Preferences via pairwise comparisons ] (appeared in AISTATS 2024). Bio: Ramya Korlakai Vinayak is the Dugald C. Jackson assistant professor in the Dept. of ECE and affiliated faculty in the Dept. of Computer Science and the Dept. of Statistics at the UW-Madison. Her research interests span the areas of machine learning, statistical inference, and crowdsourcing, with a focus on preference learning and alignment under heterogeneity, reliable and efficient dataset creation, and human-in-the-loop systems. Her works aim to address theoretical and practical challenges that arise when learning from heterogeneous societal data. Prior to joining UW-Madison, Ramya was a postdoctoral researcher in the Paul G. Allen School of Computer Science and Engineering at the University of Washington. She received her Ph.D. in Electrical Engineering from Caltech. She obtained her Masters from Caltech and Bachelors from IIT Madras. She is a recipient of the Schlumberger Foundation Faculty of the Future fellowship from 2013-15, and an invited participant at the Rising Stars in EECS workshop in 2019. She is the recipient of NSF CAREER Award in 2023.

1 0

Liyi Zhang will present his General Exam "Explaining Distributed Representations by Deep Neural Networks with Bayesian Inference" on Friday, October 4, 2024 at 10:00 AM in Friend Center 005 and via zoom.
by CS Grad Department 01 Oct '24

01 Oct '24

Liyi Zhang will present his General Exam " Explaining Distributed Representations by Deep Neural Networks with Bayesian Inference " on Friday, October 4, 2024 at 10:00 AM in Friend Center 005 and via zoom. Zoom link: https://princeton.zoom.us/j/9603487644 Committee Members: Tom Griffiths (advisor), Ryan Adams , Danqi Chen Abstract: Human-like intelligence in machines has been of interest to philosophers and engineers for centuries. Today’s deep learning allows computational models to obtain meaningful representations on high-dimensional inputs such as text and images. However, understanding these representations is challenging and why they work remains opaque. Our aim goes in two directions: 1) using computational foundations of human cognition to understand and to improve deep neural networks; 2) reverse-engineering the mind by using success in deep neural networks to understand human cognitive processes. In both directions, Bayesian inference serves as a bridge between the two fields. In the first goal, we present a Bayesian interpretation of the autoregressive objective under several distributional assumptions to show where the optimal content of language model embeddings can be identified. In the second goal, we show how variants of prior distributions in variational auto-encoders, a probabilistic generative model, explain how forming prototypes contributes to semi-supervised categorization. Reading List: https://docs.google.com/document/d/13_DPdcPTKklMbzzabiUXY7q-BiLmjnBJGoPZvge… 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