BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Computer Science and Engineering - ECPv6.13.0//NONSGML v1.0//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
X-ORIGINAL-URL:https://homecse.iitd.ac.in
X-WR-CALDESC:Events for Computer Science and Engineering
REFRESH-INTERVAL;VALUE=DURATION:PT1H
X-Robots-Tag:noindex
X-PUBLISHED-TTL:PT1H
BEGIN:VTIMEZONE
TZID:Asia/Kolkata
BEGIN:STANDARD
TZOFFSETFROM:+0530
TZOFFSETTO:+0530
TZNAME:IST
DTSTART:20250101T000000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTART;TZID=Asia/Kolkata:20260825T120000
DTEND;TZID=Asia/Kolkata:20260825T130000
DTSTAMP:20260921T143430
CREATED:20260818T060107Z
LAST-MODIFIED:20260818T060107Z
UID:2651-1787659200-1787662800@homecse.iitd.ac.in
SUMMARY:Scaling Up GPU Memory Management by Dr. Pratheek B
DESCRIPTION:Abstract:  \nThe volume of data generated worldwide is growing at an unprecedented rate\, and GPUs have emerged as the primary compute engine for processing this data. While GPUs offer massive compute power — reaching thousands of TFLOPS — they are constrained by the relatively low memory bandwidth (around a few TB/s) and limited memory capacity (in tens of GBs). As a result\, memory is often the primary performance bottleneck in many GPU applications.\n\nIn this talk\, we explore two key aspects of GPU memory management: memory oversubscription and address translation. Memory oversubscription enables GPU programs to work on datasets larger than the on-board GPU memory\, but can lead to severe slowdowns. Efficient address translation is important in GPUs\, as it lies in the critical path of memory accesses\, and thus impacts overall GPU performance.\n\nThis talk will primarily focus on the challenges posed by GPU memory oversubscription. GPU memory oversubscription enables GPU applications to work with datasets larger than the GPU memory capacity\, using the CPU memory as swap space. Unfortunately\, applications under GPU memory oversubscription often experience significant slowdowns. Our work\, ObservUVM\, improves GPU's eviction and prefetching policies by enabling observability into GPU’s memory accesses to pages resident on GPU memory — something current GPUs lack. We show that  current eviction and prefetching policies\, handled by the driver running on the CPU\, are limited in their ability to make informed decisions due to the lack of observability. ObservUVM enables observability into GPU’s memory accesses by repurposing existing hardware access counters\, enabling better-informed eviction and prefetching policies. ObservUVM improves UVM performance by around 33% across 14\napplications.\n\nWe will also touch upon a few other relevant problems in GPU memory management. We will discuss SUV\, our compiler-driven technique to perform automated data placement and migration in memory-oversubscribed GPUs. Then we will briefly discuss the impact of non-uniformity of Multi-Chip-Module (MCM) design on address\ntranslation in GPUs\, and on the impact of multi-tenancy on address translation in GPUs.\n\n\nBio:  \nPratheek is a Senior Software Engineer at NVIDIA\, working on improving GPU memory management. He obtained his Ph.D. from the Indian Institute of Science\, Bengaluru in 2026. His research revolves around improving GPU memory management\, focusing on address translation and data placement for large-memory workloads. His work spans the domains of GPU micro-architecture\, compiler techniques\, and operating systems.\nPreviously\, he had worked on reverse-engineering Nvidia GPUs.
URL:https://homecse.iitd.ac.in/event/scaling-up-gpu-memory-management-by-dr-pratheek-b/
LOCATION:Bharti 501\, IIT Campus\, Hauz Khas\, New Delhi
CATEGORIES:Seminars
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Kolkata:20260819T110000
DTEND;TZID=Asia/Kolkata:20260819T120000
DTSTAMP:20260921T143430
CREATED:20260813T065722Z
LAST-MODIFIED:20260813T065722Z
UID:2636-1787137200-1787140800@homecse.iitd.ac.in
SUMMARY:Foundations of Learning from Positive Samples by Dr. Anay Mehrotra
DESCRIPTION:Abstract: What can be learned from data? Traditional answers to this question assume an idealized data-generating process where test and training distributions are symmetric\, which is rarely the case in applications.\nIn this talk\, we will revisit this question for positive-only learning\, a setting where only positive examples are observed. This is a challenging problem arising in bioinformatics and causal inference. Classical results show that learning is impossible in general. However\, we will show that the hard instances are fragile: under a smoothed analysis\, which rules out pathological distributions\, efficient learning is possible. The same ideas also lead to faster and more general algorithms for estimation and regression from truncated data\, a foundational problem in statistics.\nThe talk is based on joint work with Shai Ben-David\, Yang Cai\, Constantine Caramanis\, Alkis Kalavasis\, Alex Kouridakis\, Jane H. Lee\, Katerina Mamali\, Farnam Mansouri\, and Manolis Zampetakis. \nSpeaker bio: Anay Mehrotra (https://anaymehrotra.com) is a Motwani Postdoctoral Fellow at Stanford working with Amin Saberi. He recently completed his PhD at Yale\, advised by Amin Karbasi and Manolis Zampetakis. His work has received the Best Paper Award at COLT and the Sri Binay Kumar Sinha Award from IIT Kanpur\, was selected for the TCS4All Rising Star session at STOC 2026\, and has been featured in WIRED.
URL:https://homecse.iitd.ac.in/event/foundations-of-learning-from-positive-samples-by-dr-anay-mehrotra/
LOCATION:Bharti 501\, IIT Campus\, Hauz Khas\, New Delhi
CATEGORIES:Seminars
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Kolkata:20260814T140000
DTEND;TZID=Asia/Kolkata:20260814T150000
DTSTAMP:20260921T143430
CREATED:20260813T064616Z
LAST-MODIFIED:20260813T064616Z
UID:2628-1786716000-1786719600@homecse.iitd.ac.in
SUMMARY:Weight Enumerators and Magic State Distillation by Dr. Amolak Kalra
DESCRIPTION:In this talk I will start by describing a protocol called magic state distillation\, which was first introduced by Bravyi and Kitaev in 2005. This protocol uses quantum error-correcting codes and noisy magic states to perform universal fault-tolerant quantum computation.\nI will then explain how the performance of a certain class of magic state distillation protocols can be determined by the weight enumerator of the quantum error-correcting code used for distillation. Using this connection\, I will show how one can derive new constraints on certain families of classical and quantum weight enumerators that are stronger than those previously known.\nThese stronger constraints lead to tighter distance bounds for quantum codes and\, rather surprisingly\, to a proof of the non-existence of certain families of classical codes.\nThis talk is based on joint work with Shiroman Prakash: https://arxiv.org/abs/2501.10163
URL:https://homecse.iitd.ac.in/event/weight-enumerators-and-magic-state-distillation-by-dr-amolak-kalra/
LOCATION:Bharti 501\, IIT Campus\, Hauz Khas\, New Delhi
CATEGORIES:Seminars
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Kolkata:20260814T120000
DTEND;TZID=Asia/Kolkata:20260814T130000
DTSTAMP:20260921T143430
CREATED:20260813T064455Z
LAST-MODIFIED:20260813T064455Z
UID:2626-1786708800-1786712400@homecse.iitd.ac.in
SUMMARY:Two New Frontiers in Science: Autonomous AI Scientists and Lotteries for Acceptance Decisions by Prof Nihar Shah
DESCRIPTION:Abstract:\nWe will discuss two new frontiers in science. \n1) Autonomous AI Scientists: Research conducted by autonomous AI scientist systems is rapidly increasing in prevalence. These systems execute the entire research process autonomously\, with little or no human intervention. While papers generated may appear appealing\, we investigate whether these systems follow rigorous research methodologies. Using novel experiments designed to mitigate confounding factors\, we uncover significant and concerning methodological flaws in their workflows. We also propose a method to detect these problems and provide policy recommendations for journals and conferences: Such methodological problems are not detectable from the produced paper alone but can be identified through analysis of the trace logs of the workflow executed by the AI scientists. \n2) Lotteries for Acceptance Decisions: Traditional decisions for accepting/rejecting papers or grant proposals involve expert reviews followed by discussions and human-specified decisions. More recently though\, citing drawbacks of such traditional approaches\, a number of funding agencies worldwide have moved towards a different decision model to make acceptance decisions. These agencies have incorporated “partial lotteries” into their decision-making\, where final decisions are randomized in a manner that still respects reviewers’ evaluations. We will first identify several problems in current implementations of such partial lotteries. We will then present a principled approach to designing improved partial lotteries with strong mathematical guarantees and empirical performance. \nThe talk will also contain a generous dose of minions. \n  \nBio: Nihar B. Shah is an Associate Professor in the Machine Learning and Computer Science departments at Carnegie Mellon University (CMU). His research focuses on the Evaluation of Science and the Science of Evaluation. His group develops computational tools with strong theoretical guarantees\, and designs and conducts controlled experiments for evidence-based policy design. His work has been used in the review of well over a hundred thousand papers and thousands of proposals\, across over 200 venues. He is a recipient of The Allen Newell Award for Research Excellence\, a Young Alumnus Medal from the Indian Institute of Science\, a JP Morgan faculty research award\, Google Research Scholar Award\, an NSF CAREER Award\, and the David J. Sakrison memorial prize from EECS Berkeley for a “truly outstanding and innovative PhD thesis.” Papers authored by him have won several Best Paper Awards.
URL:https://homecse.iitd.ac.in/event/two-new-frontiers-in-science-autonomous-ai-scientists-and-lotteries-for-acceptance-decisions-by-prof-nihar-shah/
LOCATION:Bharti 501\, IIT Campus\, Hauz Khas\, New Delhi
CATEGORIES:Seminars
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Kolkata:20260813T120000
DTEND;TZID=Asia/Kolkata:20260813T130000
DTSTAMP:20260921T143430
CREATED:20260813T050927Z
LAST-MODIFIED:20260813T050927Z
UID:2622-1786622400-1786626000@homecse.iitd.ac.in
SUMMARY:Abstractions for expressive\, extensible\, and scalable root cause analysis by Dr. Vipul Harsh
DESCRIPTION:Abstract: Modern Internet-scale services must identify and mitigate customer-impacting incidents quickly. Despite the development of many Root Cause Analysis (RCA) algorithms—including recent LLM-assisted solutions—existing approaches struggle with the “long tail” of novel failure modes and the sheer scale of telemetry. In this talk\, I argue that the path forward requires a paradigm shift: from developing point-solution algorithms to a systems-first approach that enables rapid prototyping of diverse RCA techniques. I will introduce MoCE: a Mixture-of-Experts framework that provides high-level abstractions for failure diagnosis. MoCE lets developers express complex troubleshooting logic succinctly via its domain-specific operators while its underlying runtime system handles scalable telemetry processing. Finally\, I will briefly describe how these abstractions enable reliable\, autonomous agents for domain-specific data analytics tasks.\n\nBio: Vipul Harsh is a postdoctoral researcher at Conviva with Vyas Sekar and a visiting researcher at Carnegie Mellon University (CMU). His research interests lie at the intersection of networked systems\, AI\, and theory. His current research focuses on developing new abstractions and frameworks for building reliable and verifiable systems — from networked infrastructure to AI agents. He has also worked on datacenter topology\, distributed monitoring\, and parallel algorithms. His research has been published in top-tier CS conferences (SIGCOMM\, NSDI\, SPAA\, among others); his work on Murphy was adopted into VMware’s network management suite and DRing/Starfish have directly influenced Amazon AWS’s datacenter architecture. He completed his Ph.D. at UIUC\, where he worked with Brighten Godfrey\, and holds an undergraduate degree from IIT Bombay. His thesis was nominated by UIUC for the ACM SIGCOMM dissertation award.
URL:https://homecse.iitd.ac.in/event/abstractions-for-expressive-extensible-and-scalable-root-cause-analysis-by-dr-vipul-harsh/
CATEGORIES:Seminars
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Kolkata:20260812T120000
DTEND;TZID=Asia/Kolkata:20260812T130000
DTSTAMP:20260921T143430
CREATED:20260813T051313Z
LAST-MODIFIED:20260813T051313Z
UID:2624-1786536000-1786539600@homecse.iitd.ac.in
SUMMARY:Building Trustworthy Intelligent Systems for Critical Infrastructures by Dr. Geetanjali
DESCRIPTION:Abstract: The convergence of artificial intelligence (AI)\, the Internet of Things (IoT)\, cloud-edge computing\, and autonomous cyber-physical systems is transforming critical infrastructures across healthcare\, manufacturing\, transportation\, and smart cities. As these technologies become increasingly interconnected\, ensuring that they operate in a secure\, reliable\, explainable\, and privacy-preserving manner has emerged as a fundamental research challenge. This talk presents a coherent research program dedicated to the development of Trustworthy Intelligent Systems through the integration of network security\, blockchain\, trust management\, intelligent decision-making\, and AI. \nIt traces the evolution of this research from blockchain-enabled secure communication and privacy-preserving healthcare frameworks to adaptive trust evaluation\, multi-criteria decision-making\, and decentralized trust management for the Industrial Internet of Things (IIoT). These contributions culminate in TrustNextGen\, a comprehensive framework for trustworthy next-generation IIoT\, providing a robust foundation for future advances in Trustworthy AI. The talk concludes by outlining a forward-looking research roadmap encompassing explainable AI\, federated learning\, edge intelligence\, digital twins\, and resilient autonomous systems. Together\, these directions aim to enable secure\, trustworthy\, and human-centric intelligent systems capable of safeguarding next-generation critical infrastructures while delivering broad societal and industrial impact. \nBio: Dr. Geetanjali Rathee is an Assistant Professor in the Department of Computer Science andEngineering at Netaji Subhas University of Technology (NSUT)\, Dwarka\, New Delhi. Prior tojoining NSUT\, she served as an Assistant Professor (Senior Grade) at Jaypee University\nofInformation Technology (JUIT)\, Waknaghat\, Himachal Pradesh\, for four years. She earned herB.Tech.\, M.Tech.\, and Ph.D. degrees in Computer\nScience and Engineering in 2011\, 2014\, and2017\, respectively.Dr.Rathee has made significant research contributions in the fields of\ncomputer science andnetwork security. She holds eight Indian patents and has authored approximately 15 papers inIEEE Transactions journals\n(including publications in journals with impact factors as high as9.1)\, over 40 SCI-indexed research papers\, around 10 Scopus-indexed\npublications\, and morethan 15 conference papers and book chapters at national and international forums.Her research has also been supported\nthrough competitive funding. She successfully completeda SERB-SURE (ANRF) sponsored project on data governance and security in\nhealthcare\, whileanother collaborative proposal with Sir Ganga Ram Hospital\, Delhi\, is currently under reviewunder the ICRM scheme. In\naddition\, Dr. Rathee is the author of the book Large-Scale DataStreaming\, Processing\, and Blockchain Security. She has delivered invited talks on blockchaintechnology and network security at several reputed universities and institutions across India.Her research technology\,resilient wireless mesh networks\, routing protocols\, computer networking\, and Industry 4.0.She also serves as a regular reviewer for several leading international journals\, including IEEETransactions on Vehicular Technology\, Wireless Networks\, Cluster Computing\, Journal ofAmbient Intelligence and Humanized Computing\, Transactions on Emerging Telecommunications Technologies\, and the International Journal of Communication Systems.
URL:https://homecse.iitd.ac.in/event/building-trustworthy-intelligent-systems-for-critical-infrastructures-by-dr-geetanjali/
CATEGORIES:Seminars
ORGANIZER;CN="Subodh Sharma":MAILTO:svs@cse.iitd.ac.in
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Kolkata:20260810T120000
DTEND;TZID=Asia/Kolkata:20260810T130000
DTSTAMP:20260921T143430
CREATED:20260813T064721Z
LAST-MODIFIED:20260813T064721Z
UID:2630-1786363200-1786366800@homecse.iitd.ac.in
SUMMARY:Cryptographic proofs for privacy and integrity by Prof. Chaya Ganesh (IISc)
DESCRIPTION:Abstract: \nA common denominator of conventional financial systems\, trusted execution environments (like SGX)\, blockchain technology\, and ZK-rollups is the promise of computational integrity — doing the right computation on potentially secret inputs\, even when there is no trust. \n  \nIn this talk\, we will define computational integrity and show how one can verify the correctness of a computation much more efficiently than having to re-perform the computation. We will introduce the notion of succinct proof systems that allow a prover to convince a verifier about the correctness of computation such that verification is exponentially faster than the computation itself\, and zero-knowledge where the verifier learns nothing beyond the truth of the statement. We will see applications of zkSNARKs (Zero-knowledge Succinct Non-interactive ARguments of Knowledge)\, a kind of succinct arguments in decentralized systems like blockchain technology for both privacy and scalability issues\, and outline the design principle underlying zkSNARK constructions.
URL:https://homecse.iitd.ac.in/event/cryptographic-proofs-for-privacy-and-integrity-by-prof-chaya-ganesh-iisc/
LOCATION:Bharti 501\, IIT Campus\, Hauz Khas\, New Delhi
CATEGORIES:Seminars
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Kolkata:20260805T110000
DTEND;TZID=Asia/Kolkata:20260805T120000
DTSTAMP:20260921T143430
CREATED:20260813T064916Z
LAST-MODIFIED:20260813T064916Z
UID:2632-1785927600-1785931200@homecse.iitd.ac.in
SUMMARY:Finding Equal Subset Sums in the Pigeonhole Regime by Dr. Pranjal Dutta
DESCRIPTION:Abstract: The Pigeonhole Equal Subset Sum problem (PESS)\, introduced by Papadimitriou (1994)\, asks: given n positive integers bounded by M with total sum less than 2^n − 1\, find two distinct subsets with the same sum. A solution is guaranteed by the pigeonhole principle\, yet finding one efficiently has been a longstanding challenge.\nIn this talk\, I will introduce the problem and discuss its connections to Subset Sum\, Equal Subset Sum\, and total search problems. First\, I will describe a simple birthday-paradox-based algorithm for the weak-pigeonhole regime\, and explain how combining it with Karmarkar–Karp differencing yields faster algorithms for dense instances. Second\, I will discuss a deterministic poly(n) · M^{o(1)}-time algorithm when M = 2^{o(n)}\, based on block merging and modular pruning. I will also discuss a conditional lower bound from lattice problems\, as well as an average-case poly(n) · M^{1/4}-time algorithm. This beats the best known algorithm which runs in poly(n)·M^{1/3} time (Jin-Wu\, ICALP 2024\, Jin-Williams-Zhang\, ESA 2025).\nBased on joint work with Deepak Bhati\, Antoine Joux\, Mahesh Sreekumar Rajasree\, and Karol Węgrzycki\, which got accepted in FOCS 2026. \nBio: Pranjal currently holds the Nanyang Assistant Professorship in the College of Computing and Data Science (CCDS) at Nanyang Technological University (NTU) Singapore. He spent Fall 2025 at Simons Institute as a Simons-Berkeley Fellow as well as Jane Street Research Fellow. Before joining NTU\, he was a postdoc at NUS Singapore\, hosted by Prof. Divesh Aggarwal. He obtained his PhD from CMI\, advised by Prof. Nitin Saxena and he was supported by Google PhD Fellowship. His PhD work won the ACM India Doctoral Dissertation Award 2023. He is broadly interested in Theoretical Computer Science\, with focus on algberaic flavoured algorithmic questions.
URL:https://homecse.iitd.ac.in/event/finding-equal-subset-sums-in-the-pigeonhole-regime-by-dr-pranjal-dutta/
LOCATION:Bharti 501\, IIT Campus\, Hauz Khas\, New Delhi
CATEGORIES:Seminars
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Kolkata:20260727T110000
DTEND;TZID=Asia/Kolkata:20260727T120000
DTSTAMP:20260921T143430
CREATED:20260813T065116Z
LAST-MODIFIED:20260813T065116Z
UID:2634-1785150000-1785153600@homecse.iitd.ac.in
SUMMARY:The Mathematics (and Ethics) of Fair Resource Allocation by Dr. Siddhartha Banerjee (Cornell University)
DESCRIPTION:Abstract: In many settings\, a finite supply of some public resource is allocated among people over time\, without using money: a computing cluster among researchers\, food among food-banks\, medical supplies between hospitals\, funding between non-profit projects\, fellowships among admitted students\, etc. The underlying aim is often to try and be ‘fair’ in these allocations…but what exactly do we mean? \nUnderstanding fairness in sequential decision-making is one of the most urgent (but also\, intellectually beautiful) topics today\, with deep connections to control theory\, economics\, optimization\, and normative philosophy. I will demonstrate this using two case studies: one where we know the exact numerical utility each agent gets from being allocated; and one where we do not\, but still want some rules as to which agents should have priority over others\, and why. Through these examples\, I will try to convince you of one big idea: as applied mathematicians\, our main task should be to characterize trade-offs between fairness and efficiency in different settings\, as a guide to policy-makers to help determine what is socially relevant. Luckily\, doing so turns out to be very intellectually rewarding — I will try and describe surprising connections between our problems and a host of other topics — physics\, information theory\, combinatorics\, computational complexity\, and many more! \nBio:  Sid Banerjee is an associate professor in the School of Operations Research at Cornell\, working on topics at the intersection of data-driven decision-making\, network algorithms and market design. His research is supported by grants from the NSF (including an NSF CAREER award)\, ARO\, and AFOSR\, and has received multiple awards including the Erlang Prize (2022)\, and best paper awards at ACM SIGMETRICS (2026) and the INFORMS Applied Probability Society (2021). He completed his PhD from the ECE Department at UT Austin\, and was a postdoctoral researcher in the Social Algorithms Lab at Stanford. He also served as a technical consultant with the research science group at Lyft from 2014-18.
URL:https://homecse.iitd.ac.in/event/the-mathematics-and-ethics-of-fair-resource-allocation-by-dr-siddhartha-banerjee-cornell-university/
LOCATION:Bharti 501\, IIT Campus\, Hauz Khas\, New Delhi
CATEGORIES:Seminars
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Kolkata:20260528T110000
DTEND;TZID=Asia/Kolkata:20260528T120000
DTSTAMP:20260921T143430
CREATED:20260515T060527Z
LAST-MODIFIED:20260515T060527Z
UID:2523-1779966000-1779969600@homecse.iitd.ac.in
SUMMARY:Faster Queries\, Smarter Execution: Factorization Meets Vectorization in Modern Data Systems
DESCRIPTION:Abstract:\n\nThis talk explores a novel approach to speeding up complex database queries by combining two powerful ideas: factorization and vectorized execution. Instead of processing large intermediate results (which can be slow and memory-intensive)\, the method represents data in a compact\, factorized form that avoids redundancy. It also applies vectorized processing techniques to process data efficiently at the hardware level.\n\nSpeaker Bio:\nSunny Yasser is a PhD student in Computer Engineering at Polytechnique Montréal\, in the Data and AI Systems (DAIS) Lab under Prof. Amine Mhedhbi. He is also associated with MILA\, the premier AI research lab started by Prof. Yoshua Bengio. His research centers on high-performance query processing\, with an emphasis on compression-aware execution on modern hardware.  His broader work explores low-level system optimizations for scalable analytical and AI-driven data workloads. He has published papers at VLDB and SIGMOD previously.
URL:https://homecse.iitd.ac.in/event/faster-queries-smarter-execution-factorization-meets-vectorization-in-modern-data-systems/
LOCATION:Bharti 501\, IIT Campus\, Hauz Khas\, New Delhi
CATEGORIES:Seminars
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Kolkata:20260306T120000
DTEND;TZID=Asia/Kolkata:20260306T130000
DTSTAMP:20260921T143430
CREATED:20260303T103646Z
LAST-MODIFIED:20260303T103646Z
UID:2432-1772798400-1772802000@homecse.iitd.ac.in
SUMMARY:Neural Circuit Discovery via Representation and Dynamics by Savik Kinger
DESCRIPTION:Abstract: Neuroscience and AI share a bottleneck: while one can build (artificial) or record (biological) complex networks\, we struggle to explain their functional circuitry; i.e.\, how they compute. In this talk I use whole-brain recordings from C. elegans\, a canonical neurobiological system\, as a concrete testbed for “circuit interpretability.” I then introduce two complementary inference approaches for turning high-dimensional activity data into mechanistic structure. Approach 1 treats circuit discovery as a representation problem: learn time-varying functional structure and uncover recurring\, stimulus-dependent modules rather than a single static connectivity map. Approach 2 treats circuit discovery as a dynamics problem: go beyond correlation to estimate directed\, time-lagged influence—i.e.\, which units appear to drive others and over what delays—using modern score-based generative modeling ideas. I will show how these ML methods produce testable hypotheses for biologists and\, potentially\, offer new avenues for understanding complex networks in AI. \nBio: Savik Kinger is a PhD candidate in Computer Science at Yale University\, advised by Steven Zucker. His research focuses on developing methods to analyze biological and artificial neural networks\, integrating ideas from machine learning\, dynamical systems\, and causal inference. He received Bachelor’s degrees in Math and Computer Science from Columbia University. He is supported by a Nathan Hale fellowship.
URL:https://homecse.iitd.ac.in/event/neural-circuit-discovery-via-representation-and-dynamics-by-savik-kinger/
LOCATION:SIT 113\, Amar Nath and Shashi Khosla School of Information Technology\, Indian Institute of Technology\, Delhi\, Hauz Khas\, New Delhi\, Delhi\, 110016\, India
CATEGORIES:Seminars
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Kolkata:20260202T120000
DTEND;TZID=Asia/Kolkata:20260202T130000
DTSTAMP:20260921T143430
CREATED:20260115T072749Z
LAST-MODIFIED:20260115T072749Z
UID:2365-1770033600-1770037200@homecse.iitd.ac.in
SUMMARY:Deep generative models for single-cell and spatial genomics by Ajita Shree
DESCRIPTION:Speaker: Ms. Ajita Shree is a PhD student in the Department of Computer Science and Engineering at IIT Kanpur and will be joining EMBL-EBI\, UK\, as a postdoctoral researcher. \nAbstract: Recent advances in large-scale genomic assays\, including single-cell and spatial transcriptomics (ST)\, have provided unprecedented insights into the biological mechanisms underlying development\, disease\, and therapeutic response. However\, these datasets pose significant computational challenges. \nOne of the major challenges is the integration of heterogeneous single-cell datasets across donors\, time points\, and experimental conditions to create a unified resource for downstream analysis. In this talk\, I will present scDREAMER\, a novel deep generative model that can perform integration of multi-batch single-cell datasets in unsupervised\, semi-supervised and supervised settings. Using real benchmarking datasets\, we demonstrate that scDREAMER can overcome critical challenges including skewed cell type distribution among batches\, nested batch-effects\, large number of batches and enables integration of millions of cells across species. \nFurther\, I will discuss the emerging field of spatial genomics\, which enables the study of gene expression at a spatial level\, but its spot-level resolution poses challenges in resolving cell-type contributions in situ. This has spurred extensive development of deconvolution methods\, yet evaluation remains lacking. In this work\, we present a novel graph attention auto-encoder for simulating spatial cell type distributions from three major tissue types including brain\, cancer and organs\, and performed a comprehensive benchmarking. \nBio: Ajita Shree is a PhD student in the Department of Computer Science and Engineering at IIT Kanpur and will be joining EMBL-EBI\, UK\, as a postdoctoral researcher. Previously\, she worked for three years as a Data Scientist at GE Aerospace and GE Global Research\, and was a graduate of GE’s Global Leadership Program. She earned her M.Tech. in Computer Science and Engineering from IIT Delhi in 2017 and holds a B.Tech. in Computer Science and Engineering.
URL:https://homecse.iitd.ac.in/event/deep-generative-models-for-single-cell-and-spatial-genomics-by-ajita-shree/
LOCATION:Bharti 501\, IIT Campus\, Hauz Khas\, New Delhi
CATEGORIES:Seminars
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Kolkata:20260122T120000
DTEND;TZID=Asia/Kolkata:20260122T130000
DTSTAMP:20260921T143430
CREATED:20260116T114035Z
LAST-MODIFIED:20260116T144924Z
UID:2385-1769083200-1769086800@homecse.iitd.ac.in
SUMMARY:Logical Relations for Formally Verified Authenticated Data Structures by Chaitanya Agarwal
DESCRIPTION:Venue: Bharti501 \nAbstract: Authenticated data structures (ADSs) allow untrusted third parties to carry out operations which produce proofs that can be used to verify an operation’s output. Such data structures are challenging to develop and implement correctly. In this talk\, I will talk about a library\, Authentikit\, that is implemented in OCaml\, that generates authenticated versions of data structures automatically. I will also talk about recent work by us (https://dl.acm.org/doi/abs/10.1145/3719027.3744801) that gives a formal proof of security and correctness of Authentikit. The proof is based on a new relational separation logic for reasoning about programs that use collision-resistant cryptographic hash functions. This logic provides a basis for constructing two semantic models of a type system\, which are used to justify how Authentikit makes use of type abstraction to enforce security and correctness. Using these models we also prove the correctness of several optimizations to Authentikit and then show how optimized\, hand-written implementations of authenticated data structures can be soundly linked with automatically generated code. All of the results have been mechanized in the Rocq prover using the Iris framework. \nSpeaker Bio: Chaitanya Agarwal (https://culechetoo.github.io <https://culechetoo.github.io/>) is a 3rd year computer science PhD student at the New York University\, advised by Joseph Tassarotti. He is broadly interested in programming languages and formal verification with a particular focus on verification of security applications. In the past\, he has also worked with Thomas Wies on abstract-interpretation analysis for recursive\, higher-order programs\, and with Shibashis Guha\, on developing statistical-model-checking techniques for Markov Decision Processes (MDPs). Chaitanya obtained his B.Tech. from IIIT Delhi.
URL:https://homecse.iitd.ac.in/event/logical-relations-for-formally-verified-authenticated-data-structures-by-chaitanya-agarwal/
LOCATION:Bharti 501\, IIT Campus\, Hauz Khas\, New Delhi
CATEGORIES:Seminars
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Kolkata:20251215T120000
DTEND;TZID=Asia/Kolkata:20251215T130000
DTSTAMP:20260921T143430
CREATED:20251209T045950Z
LAST-MODIFIED:20251209T045950Z
UID:2235-1765800000-1765803600@homecse.iitd.ac.in
SUMMARY:Matroids are Equitable by Hannaneh Akrami
DESCRIPTION:Abstract: We show that if the ground set of a matroid can be partitioned into k≥2 bases\, then for any given subset S of the ground set\, there is a partition into k bases such that the sizes of the intersections of the bases with S may differ by at most one. This settles the matroid equitability conjecture by Fekete and Szabó (Electron.~J.~Comb.~2011) in the affirmative. We also investigate equitable splittings of two disjoint sets S1 and S2\, and show that there is a partition into k bases such that the sizes of the intersections with S1 may differ by at most one and the sizes of the intersections with S2 may differ by at most two; this is the best possible one can hope for arbitrary matroids. \nWe also derive applications of this result into matroid constrained fair division problems. We show that there exists a matroid-constrained fair division that is envy-free up to 1 item if the valuations are identical and tri-valued additive. We also show that for bi-valued additive valuations\, there exists a matroid-constrained allocation that provides everyone their maximin share. \nThis is based on joint work with Siyue Liu\, Roshan Raj\, and László A. Végh. \nSpeaker Bio: Hannaneh Akrami is a postdoctoral fellow at the University of Bonn and a Minerva Fast-track fellow at MPI-Informatik. She obtained her PhD from the University of Saarlandes and MPI-Informatik in 2024 and completed a BSc from Sharif University of Technology in 2019. Her interests are in fair division\, Algorithmic Game theory\, Combinatorics\, Graph theory and Approximation algorithms. \n 
URL:https://homecse.iitd.ac.in/event/matroids-are-equitable-by-hannaneh-akrami/
LOCATION:Bharti 501\, IIT Campus\, Hauz Khas\, New Delhi
CATEGORIES:Seminars
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Kolkata:20251211T110000
DTEND;TZID=Asia/Kolkata:20251211T170000
DTSTAMP:20260921T143430
CREATED:20251209T113155Z
LAST-MODIFIED:20251209T113155Z
UID:2239-1765450800-1765472400@homecse.iitd.ac.in
SUMMARY:Multiparty Session Types: Separation and Encodability Results by Prof. Nobuko Yoshida
DESCRIPTION:Venue: Bharti501 / Teams link will also be shared \nAbstract: Multiparty session types (MPST) are a type discipline for enforcing the structured\, deadlock-free communication of concurrent and message-passing programs. Traditional MPST have a limited form of choice in which alternative communication possibilities are offered by a single participant and selected by another. Mixed choice multiparty session types (MCMP) extend the choice construct to include both selections and offers in the same choice. This talk presents a mixed-choice synchronous multiparty session calculus and its typing system\, which guarantees communication safety and deadlock-freedom. We then talk of expressiveness of nine subcalculi of the MCMP-calculus by examining their encodability (there exists a good encoding from one to another) and separation (there exists no good encoding from one calculus to another). The highlight is that the binary (2-party) mixed sessions by Casal et al. (2022) are strictly less expressive than the MCMP-calculus. \n\nJoint work with Kirstin Peters appeared in LICS’24 (https://arxiv.org/abs/2405.08104) \nAbout the speaker. Nobuko Yoshida is Christopher Strachey Chair of Computer Science in University of Oxford. She is an EPSRC Established Career Fellow and an Honorary Fellow at Glasgow University. Last 10 years\, her main research interests are theories and applications of protocol specification and verification. She introduced multiparty session types [ POPL’08\, JACM ] which received the Most Influential POPL Paper Award in 2018 (judged by its influence over the last decade). This work enlarged the community and widened the scope of applications of session types\, e.g. runtime monitoring based on Scribble (co-developed with Red Hat) has been deployed to other projects such as cyberinfrastructure in the US Ocean Observatories Initiative (OOI); and widened the scope of her research areas. She received the Test-of-time-award from PPDP’24 and the best paper awards from CC’20\, COORDINATION’23 and DisCoTech’23. She received the third Suffrage Science Award for Mathematics and Computing from MRC for her STEM activity. She is an editor of ACM Transactions on Programming Languages and Systems\, Theoretical Computer Science\,  ACM Formal Aspects of Computing\, Mathematical Structures in Computer Science\, Journal of Logical Algebraic Methods in Programming\, and the chief editor of The Computer-aided Verification and Concurrency Column for EATCS Bulletin.
URL:https://homecse.iitd.ac.in/event/multiparty-session-types-separation-and-encodability-results-by-prof-nobuko-yoshida/
LOCATION:Bharti 501\, IIT Campus\, Hauz Khas\, New Delhi
CATEGORIES:Seminars
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Kolkata:20251209T140000
DTEND;TZID=Asia/Kolkata:20251209T150000
DTSTAMP:20260921T143430
CREATED:20251208T052758Z
LAST-MODIFIED:20251208T052758Z
UID:2233-1765288800-1765292400@homecse.iitd.ac.in
SUMMARY:Approximating Optimal Broadcast of Files in a Hose-Model Network
DESCRIPTION:Speaker:Sukriti Gupta (PhD student)\, CSE Dept.\, IIT Delhi\nAbstract -\nWe consider the problem of file sharing among peers who are connected to \na common core network through links of differing upload and download \ncapacities\, as is the case in networks provisioned according to the hose \nmodel. The file is assumed to be divided into equal-sized chunks\, and a \npeer can start sending a “chunk” of the file to another peer only after \nit has received the entire chunk. The objective is to share a chunk\, \ninitially residing on one of the peers\, with all other peers in the \nleast time possible. Peers can simultaneously send/receive parts of a \nchunk to/from multiple peers\, subject to the upload and download \ncapacity constraints.
URL:https://homecse.iitd.ac.in/event/approximating-optimal-broadcast-of-files-in-a-hose-model-network/
LOCATION:Bharti 501\, IIT Campus\, Hauz Khas\, New Delhi
CATEGORIES:Seminars
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Kolkata:20251030T110000
DTEND;TZID=Asia/Kolkata:20251030T120000
DTSTAMP:20260921T143430
CREATED:20251028T055031Z
LAST-MODIFIED:20251028T055031Z
UID:2136-1761822000-1761825600@homecse.iitd.ac.in
SUMMARY:Towards Reliable LLM Reasoning: Coordinated Agents\, Variance-Aware Evaluation\, and Lean Inference by Prof. Akhil Arora
DESCRIPTION:Abstract: Large language models (LLMs) are increasingly deployed as reasoning engines\, yet their practical use remains constrained by three persistent challenges: achieving high-quality reasoning at low cost\, measuring performance reliably\, and ensuring efficient\, reproducible deployment. In this talk\, I will present a research agenda addressing these challenges through new methods\, benchmarks\, and systems for practical LLM reasoning. I begin with Next\, I turn to Finally\, I focus on Together\, these contributions chart a path toward LLM reasoning that is not only more powerful\, but also leaner\, more reliable\, and environmentally responsible.\n\n  \n\nBio: Akhil Arora is a Tenure-Track Assistant Professor of Computer Science at Aarhus University\, where he heads the CLAN for AI Research on Language and Networks (or “CLAN” for short). He is a fellow of the Copenhagen Center for Social Data Science (SODAS)\, an affiliate of the Pioneer Centre for AI and ELLIS\, and a formal collaborator of the Wikimedia Foundation\, the non-profit organization that manages Wikipedia and related projects. Akhil’s research lies broadly in human-centered AI with a focus on improving human knowledge-seeking\, bridging knowledge gaps\, and promoting knowledge equity on the Web. To this end\, he devises methods and tools blending techniques from NLP\, AI\, Graph ML\, and Computational social science. Recently\, his group has been devising robust\, trustworthy\, accessible\, and efficient LLM inference strategies.\nAkhil received his PhD in Computer Science from EPFL (2024) in Switzerland\, his MS from IIT Kanpur (2013)\, and his undergraduate degree from NCU Gurgaon (2010). In days of yore\, he spent close to five years in the industry working with the research labs of Xerox and American Express as a Research Scientist. His work on influence maximization has been recognized as the 8th most influential paper of SIGMOD 2017 by Paper Digest and received the 2018 ACM SIGMOD Most Reproducible Paper Award. He is a recipient of the prestigious EDIC Doctoral Fellowship\, an alumnus of the coveted Heidelberg Laureate Forum\, and a DAAD AINet fellow on human-centered AI. Akhil is a director of the P1-programs on
URL:https://homecse.iitd.ac.in/event/towards-reliable-llm-reasoning-coordinated-agents-variance-aware-evaluation-and-lean-inference-by-prof-akhil-arora/
LOCATION:Bharti 501\, IIT Campus\, Hauz Khas\, New Delhi
CATEGORIES:Seminars
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Kolkata:20250901T120000
DTEND;TZID=Asia/Kolkata:20250901T130000
DTSTAMP:20260921T143430
CREATED:20250806T062326Z
LAST-MODIFIED:20250809T094539Z
UID:1766-1756728000-1756731600@homecse.iitd.ac.in
SUMMARY:Giving Some Space Can Be Hard: Two New Models to Match Agents with Locations by Shivika Narang
DESCRIPTION:Abstract: There can be a multitude of reasons to match agents to specific locations in a given space. In this talk we cover two: distributing delivery orders and assigning shared hostel rooms. For both settings we shall try to find solutions that satisfy desirable properties and characterize instances for which they exist.\n \nWe first initiate the study of fair distribution of delivery tasks among a set of agents wherein delivery jobs are placed along the vertices of a graph. Our goal is to fairly distribute delivery costs (modeled as a submodular function) among a fixed set of agents while satisfying some desirable notions of economic efficiency. We characterize instances that admit fair and efficient solutions by exploiting underlying graph structures. Unfortunately\, finding these solutions proves to be NP-hard. We complement this by designing an XP algorithm (parameterized by the number of agents) that can find all fair and efficient solutions when they exist. We conclude this discussion by theoretically and experimentally analyzing the price of fairness.\n \nWe shall then introduce Leontief utilities to the problem of roommate matchings. We aim to find strategyproof mechanisms that give good bounds on agent welfare. We first find that no approximation to welfare can be achieved under strategyproof mechanisms for either Leontief or additive utilities. Even for binary additive utilities no maximum welfare mechanism can be strategyproof. In contrast\, we then –surprisingly– find that binary Leontief utilities enable us to find strategyproof mechanisms that maximize welfare.\n \nJoint work with Hadi Hosseini\, Sanjukta Roy and Tomasz Was.\n \nBio: Shivika Narang is a postdoctoral fellow at UNSW Sydney. Previously she was a postdoc at Simons Laufer Mathematical Sciences Institute\, Berkeley (SLMath) and completed her PhD from IISc Bengaluru. During her PhD\, she received the Tata Consultancy Services Research Fellowship. Her work is currently focused on finding fair and efficient solutions to societal problems. She largely works in computational social choice\, especially matching and allocation problems.
URL:https://homecse.iitd.ac.in/event/giving-some-space-can-be-hard-two-new-models-to-match-agents-with-locations-by-dr-shivika-narang/
LOCATION:Bharti 501\, IIT Campus\, Hauz Khas\, New Delhi
CATEGORIES:Seminars
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Kolkata:20250822T140000
DTEND;TZID=Asia/Kolkata:20250822T150000
DTSTAMP:20260921T143430
CREATED:20250827T082004Z
LAST-MODIFIED:20250827T082004Z
UID:1825-1755871200-1755874800@homecse.iitd.ac.in
SUMMARY:Optimal Capacity Modification for Stable Matchings with Ties by Dr. Keshav Ranjan
DESCRIPTION:Abstract: In this talk\, we consider the Hospitals/Residents (HR) problem in the presence of ties in preference lists of hospitals. Among the three notions of stability\, viz. weak\, strong\, and super stability\, we focus on strong stability. Strong stability is appealing both theoretically and practically; however\, its existence is not guaranteed. Our objective is to optimally increase hospitals’ quotas so that the resulting instance admits a strongly stable matching.\nSuch an augmentation is guaranteed to exist when resident preference lists are strict. We explore two natural optimization criteria:\n\n\n\nMINSUM: minimizing the total capacity increase across all hospitals and \nMINMAX: minimizing the maximum capacity increase for any hospital\n\nWe prove that the MINSUM problem admits a polynomial-time algorithm\, whereas the MINMAX problem is NP-hard. We prove an analogue of the Rural Hospitals theorem for the MINSUM problem. When each hospital incurs a cost for a unit increase in its quota\, the MINSUM problem becomes NP-hard\, even for 0/1 costs. In fact\, we show that the problem cannot be approximated to any multiplicative factor. We also present a polynomial-time algorithm for optimal MINSUM augmentation when a specified subset of edges is required to be included in the matching.\n\nThe talk is based on a recent work accepted at IJCAI 2025 and is a joint work with Meghana Nasre (IIT-M) and Prajakta Nimbhorkar (CMI). \nBio: Keshav Ranjan recently (July 2025) completed his Ph.D. from the Department of Computer Science and Engineering\, IIT Madras\, under the supervision of Dr. Meghana Nasre. His Doctoral thesis\, titled “Two-Sided Matchings: Lower Quotas\, Ties\, and Capacity Augmentation”\, focuses on the algorithmic aspects of two-sided matching problems under various constraints. Previously\, he held an M. Tech degree in Mathematics and Computing from the Department of Mathematics\, IIT Patna. His research interests lie in the broad area of Graph Algorithms\, with a particular focus on matching problems with preferences.
URL:https://homecse.iitd.ac.in/event/optimal-capacity-modification-for-stable-matchings-with-ties-by-dr-keshav-ranjan/
LOCATION:Bharti 501\, IIT Campus\, Hauz Khas\, New Delhi
CATEGORIES:Seminars
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Kolkata:20250806T120000
DTEND;TZID=Asia/Kolkata:20250806T130000
DTSTAMP:20260921T143430
CREATED:20250730T061826Z
LAST-MODIFIED:20250730T165814Z
UID:1727-1754481600-1754485200@homecse.iitd.ac.in
SUMMARY:Rank Aggregation and Fairness by Diptarka Chakraborty
DESCRIPTION:Abstract: Aggregating multiple input rankings over a set of candidates to generate a consensus ranking is one of the fundamental ranking problems\, having many applications in social choice theory\, hiring\, college admission\, web search\, and databases. However\, the optimal consensus ranking might be biased against any individual candidate or candidates belonging to certain marginalized communities or groups. This has motivated studies of the rank aggregation problem from the fairness perspective. While finding a consensus ranking\, the additional objective is to ensure fair representation of each group in the top positions of the final aggregated ranking. In this talk\, we will discuss various algorithms to find such a fair ranking approximately.\n\nSpeaker: Diptarka Chakraborty is an Assistant Professor at the National University of Singapore. He did his Ph.D. at the Indian Institute of Technology\, Kanpur. Before joining NUS\, he spent two years at Charles University\, Prague\, and then almost a year at Weizmann Institute of Science\, Israel\, as a post-doctoral fellow. His research interest mostly lies in theoretical computer science\, more specifically\, algorithms on large data sets\, approximation algorithms\, sublinear algorithms\, string matching algorithms\, and graph algorithms. He is a recipient of the best paper award at FOCS 2018 and the Google South & Southeast Asia Research Award 2022.
URL:https://homecse.iitd.ac.in/event/rank-aggregation-and-fairness/
LOCATION:Bharti 501\, IIT Campus\, Hauz Khas\, New Delhi
CATEGORIES:Seminars
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Kolkata:20250609T120000
DTEND;TZID=Asia/Kolkata:20250609T130000
DTSTAMP:20260921T143430
CREATED:20250604T065636Z
LAST-MODIFIED:20250604T065657Z
UID:1639-1749470400-1749474000@homecse.iitd.ac.in
SUMMARY:Designing advanced cryptographic primitives in distributed settings
DESCRIPTION:Speaker: Dr. Anshu Yadav\, Postdoctoral Researcher\, Institute of Science and Technology\, Austria \nAbstract: In today’s world\, the rapid advancement of technology has led to the generation of vast amounts of sensitive data\, which must be accessed in a secure and controlled manner to facilitate research across various domains. Often this data\, associated with a single logical entity\, is generated in a distributed manner\, yet must be protected with the same level of security as if it were produced by a single source. Furthermore\, distributing authority among multiple entities is essential to avoid a single point of security failure. These are natural\, yet complex challenges in modern cryptography. My research focuses on exploring how advanced cryptographic primitives can provide effective solutions to such problems. In this talk\, I will begin with a brief overview of my research interests and profile. I will then focus on the themes discussed above. In particular\, I will briefly talk about a result on multi-input attribute based encryption which is a generalization of attribute-based encryption (ABE) – a novel encryption paradigm enabling expressive access control on encrypted data. In ABE\, a message m is encrypted under an attribute x\, and decryption keys are associated with a policy 𝑓. Decryption is possible if and only if 𝑓(x)=1\, unlike traditional public key encryption scheme where a single key can decrypt all ciphertexts. In the multi-input setting\, data is generated by k non-interacting parties\, with each party contributing an input (x_i\, m_i)\, so that x=(x_1\,…\,x_k) and m=(m_1\,…\,m_k). The function 𝑓 is now a k-ary predicate. The goal is for each party to independently encrypt their data as ct_1\,…\,ct_k\, and for a decryption algorithm with key sk_f to recover (m_1\,…\,m_k) if and only if 𝑓(x_1\,…\,x_k)=1. We formally defined the notion of multi-input ABE (k-ABE) and presented constructions for different k under different cryptographic hardness assumptions. I will describe the key challenges in designing cryptographic schemes in the multi-input setting and how our work addresses these challenges. If time permits\, I will also briefly talk about my work in threshold cryptography – a very useful and active field of cryptography with advanced practical applications in distributed environments (e.g.\, block chains\, distributed key generation\, etc.). In threshold cryptography\, a privileged operations—such as ‘signing’ in digital signature scheme or ‘decryption’ in an encryption scheme—is distributed among n parties\, ensuring that at least a threshold t of them are required to perform the operation. In this area\, my research has mostly focussed on threshold signatures\, where we improve the security of post-quantum threshold signature scheme. Finally\, I will conclude the talk with a discussion of my future research directions\, including open problems in the areas discussed above.
URL:https://homecse.iitd.ac.in/event/designing-advanced-cryptographic-primitives-in-distributed-settings/
LOCATION:Bharti 501\, IIT Campus\, Hauz Khas\, New Delhi
CATEGORIES:Seminars
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Kolkata:20250605T100000
DTEND;TZID=Asia/Kolkata:20250605T110000
DTSTAMP:20260921T143430
CREATED:20250602T072215Z
LAST-MODIFIED:20250602T094001Z
UID:1633-1749117600-1749121200@homecse.iitd.ac.in
SUMMARY:Multimodal Learning in 3D environments: Perception\, and Simulation
DESCRIPTION:Speaker: Dr. Arun Balajee Vasudevan is currently a Research Scientist at Amazon \nAbstract: \nAutonomous robots have several potential applications such as virtual assistants\, VR/AR\, gaming\, self-driving technologies\, city planning and others. To achieve autonomy\, a robot needs to see and hear the environment\, before it can converse or navigate favorably to perform a human-desired task. Precisely\, scene understanding begins with building 3D geometry\, decoding semantics\, understanding surround objects/humans\, and planning and actions. My research talk addresses these fundamental challenges independently under two broad themes: understanding geometry and multimodal perception & data-driven simulation and navigation. \nUnder geometry and perception\, I introduce the usage of several multimodalities such as the user’s gaze\, visual sensors such as cameras or range sensors (e.g. Kinect)\, and speech/language instructions from human referrals for robot perception tasks. Further\, I delve deep into the investigation of the audio sensing modality for the task using binaural sound microphones. Secondly\, regarding the geometry\, my talk addresses one of my ongoing works about the construction of digital twins of the real world with 4D reconstruction of dynamic scenes from ground visuals of a robot. \nUnder the theme of Data-driven simulation and robot navigation. Following perception\, robots must navigate and take meaningful actions in the world. This involves broadly two aspects: wayfinding and motion planning. Earlier works address wayfinding based on directional instructions\, overlooking human aspects. I briefly talk about a new paradigm that integrates principles from cognitive science with learning-based methods to tackle the challenge of language-based wayfinding for robots in real-world outdoor environments. The second aspect is motion planning for which I propose the learning of driver behavior models for MPC-based planners to build data-driven simulators. \nLong term\, I envision to bridge the above two themes to build multimodal digital twins simulators of the real-world. This potentially helps in training and testing of planners\, VR/AR setups\, gaming\, and others. Lastly\, I also cover my future research plan in the talk. \nShort Bio: \nArun Balajee Vasudevan is currently a Research Scientist at Amazon. Previously\, he was a postdoctoral researcher at Carnegie Mellon University under Prof. Deva Ramanan till March 2025. His core research interest is in Computer Vision and Multimodal Learning. He has works in multimodal (vision\, language and sounds) perception and navigation\, 3D/4D reconstruction\, Motion Planning and improving Foundational models. He published papers predominantly in vision and machine learning conferences/journals such as CVPR\, ECCV\, ICML\, IJCV\, TPAMI\, and others. He defended his PhD under Prof. Luc Van Gool at ETH Zurich. He received his MSc in Computer Science from EPFL in 2016 and an undergraduate degree in Electrical Engineering from the Indian Institute of Technology Jodhpur in the year 2014.
URL:https://homecse.iitd.ac.in/event/multimodal-learning-in-3d-environments-perception-and-simulation/
CATEGORIES:Seminars
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Kolkata:20250529T120000
DTEND;TZID=Asia/Kolkata:20250529T130000
DTSTAMP:20260921T143430
CREATED:20250526T171008Z
LAST-MODIFIED:20250526T171008Z
UID:1628-1748520000-1748523600@homecse.iitd.ac.in
SUMMARY:Giving Some Space Can Be Hard: Two New Models to Match Agents with Locations by Shivika Narang
DESCRIPTION:Title: Giving Some Space Can Be Hard: Two New Models to Match Agents with Locations \nSpeaker: Shivika Narang (UNSW Sydney)\n\nAbstract: There can be a multitude of reasons to match agents to specific locations in a given space. In this talk\, we cover two: distributing delivery orders and assigning office spaces. For both settings\, we shall try to find solutions that satisfy desirable properties and characterize instances for which they exist.\n\nWe first initiate the study of fair distribution of delivery tasks among a set of agents\, wherein delivery jobs are placed along the vertices of a graph. Our goal is to fairly distribute delivery costs (modeled as a submodular function) among a fixed set of agents while satisfying some desirable notions of economic efficiency. We characterize instances that admit fair and efficient solutions by exploiting underlying graph structures. Unfortunately\, finding these solutions proves to be NP-hard. We complement this by designing an XP algorithm (parameterized by the number of agents) that can find all fair and efficient solutions when they exist. We conclude this discussion by theoretically and experimentally analyzing the price of fairness.\n\nWe shall then introduce and analyze distance preservation games (DPGs). In DPGs\, agents express ideal distances to other agents and need to choose locations in the unit interval while preserving their ideal distances as closely as possible. We analyze the existence and computation of location profiles that are jump stable (i.e.\, no agent can benefit by moving to another location) or welfare optimal for DPGs\, respectively.\n\nJoint Work with Hadi Hosseini and Tomasz Wąs (Fair Delivery) and Haris Aziz\, Hau Chan\, Patrick Lederer\, and Toby Walsh (DPGs).\n\nSpeaker Bio: Shivika Narang is a postdoctoral fellow at UNSW Sydney. Previously\, she was a postdoc at Simons Laufer Mathematical Sciences Institute\, Berkeley (SLMath)\, and completed her PhD from IISc Bengaluru. Her work is currently focused on fairness and efficiency in computational social choice\, especially matching and allocation problems.
URL:https://homecse.iitd.ac.in/event/giving-some-space-can-be-hard-two-new-models-to-match-agents-with-locations-by-shivika-narang/
LOCATION:Bharti 501\, IIT Campus\, Hauz Khas\, New Delhi
CATEGORIES:Seminars
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Kolkata:20250528T120000
DTEND;TZID=Asia/Kolkata:20250528T130000
DTSTAMP:20260921T143430
CREATED:20250518T062803Z
LAST-MODIFIED:20250526T171039Z
UID:1592-1748433600-1748437200@homecse.iitd.ac.in
SUMMARY:Enhancing Safety and Ethical Alignment in Large Language Models by Rima Hazra
DESCRIPTION:Speaker:  Dr. Rima Hazra \n\n\nAbstract: In this talk\, we explore cutting-edge strategies for enhancing the safety and ethical alignment of large language models (LLMs). The research spans various approaches\, including red teaming and jailbreaking techniques\, which assess and improve model robustness and ethical integrity. We delve into how instruction-centric responses\, when generated by LLMs\, can increase the likelihood of unethical output\, thereby highlighting the vulnerabilities of these AI systems. Through the introduction of frameworks like ‘Safety Arithmetic’ and ‘SafeInfer\,’ we demonstrate methods to mitigate risks by manipulating model parameters and decoding-time behaviors to foster safer interactions. The discussions also emphasize the importance of safety alignment strategies and the challenges posed by integrating new knowledge through model edits\, which can paradoxically destabilize ethical guidelines. This comprehensive examination not only sheds light on the current vulnerabilities of LLMs but also presents a pathway toward more reliable and ethically aligned AI implementations. \n\nBio: Dr. Rima Hazra is a senior postdoc at Eindhoven University of Technology (TU\e)\, Netherlands. Earlier she was a Postdoctoral Researcher at the Singapore University of Technology and Design\, working in the area of AI safety alignment\, natural language processing\, and LLM reasoning. She earned her Ph.D. from the Indian Institute of Technology\, Kharagpur\, where she explored the area of Information retrieval\, NLP and graph learning. With experience in information retrieval\, NLP and graph learning\, Dr. Hazra has published several papers in prestigious CORE A* and A conferences such as AAAI\, ACL\, EMNLP\, NAACL\, ECIR\, ECMLP PKDD and JCDL. She has also received the prestigious Microsoft Academic Partnership Grant (MAPG) and the PaliGemma Academic Program award from Google for her work in AI safety alignment.
URL:https://homecse.iitd.ac.in/event/enhancing-safety-and-ethical-alignment-in-large-language-models-by-dr-rima-hazra/
LOCATION:Bharti 501\, IIT Campus\, Hauz Khas\, New Delhi
CATEGORIES:Seminars
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Kolkata:20250523T120000
DTEND;TZID=Asia/Kolkata:20250523T130000
DTSTAMP:20260921T143430
CREATED:20250519T054519Z
LAST-MODIFIED:20250520T083950Z
UID:1599-1748001600-1748005200@homecse.iitd.ac.in
SUMMARY:Next-Generation AI-Enhanced Stream Processing
DESCRIPTION:Speaker: Dr. Manisha Luthra Agnihotri is the Deputy Head of the German Research Center for Artificial Intelligence (DFKI) in Darmstadt. \nIt is an online talk. Please write to the CSE office to get the Teams link. \nAbstract: In this talk\, I will outline my vision for next-generation\, AI-enhanced data management systems through the lens of learned stream processing. Today’s stream processing platforms demand extensive manual tuning to optimize critical decisions such as query plan selection\, operator placement\, and parallelism. My vision eliminates these labor-intensive processes by leveraging zero-shot learning to automatically derive optimal configurations\, thereby radically enhancing performance and generalisability. A key contribution of my work is a novel learned operator placement optimization provided by a novel cost model that forecasts the execution costs of streaming queries on heterogeneous hardware. Particularly in IoT environments—where diverse hardware and network conditions are the norm—our approach employs graph neural networks to predict query costs accurately\, even for unseen placements and query patterns. This approach not only overcomes the generalizability limitations of existing methods but also paves the way for more robust and adaptive cost-based optimizations for stream processing systems. I will also discuss my future research directions\, focusing on extending these AI-driven techniques to multi-modal stream processing. This work aims to redefine data management by creating systems that adapt to evolving computational needs for multiple modalities\, ultimately setting new standards for understanding data inputs and autonomy in stream processing. \nBio: Manisha Luthra Agnihotri is the Deputy Head of the German Research Center for Artificial Intelligence (DFKI) in Darmstadt and a Research Group Leader at TU Darmstadt. She co-leads the Systems AI for Decision Support group with focus of research on learned system optimizations and multimodal data management. Her work sits at the dynamic intersection of machine learning\, data systems\, and hardware\, with major contributions in learned cost-based optimization and the acceleration of query workloads via GPU and RDMA technologies. \nThroughout her academic journey\, Manisha has received several prestigious awards\, including the German national Best Ph.D. Thesis award from the GI/ITG special interest group on Communication and Distributed Systems (KuVS)\, the Athena Young Investigator Award\, the Anita Borg Faculty Scholarship\, the Zeiss Top Dissertation Scholarship\, and mentoring and networking accolades from the German Research Foundation (DFG). Her expertise has led her to speak at top-tier institutions such as the University of Toronto\, and she has presented her innovative research at premier conferences like SIGMOD\, VLDB\, ICDE\, and EDBT. Manisha also actively contributes to the academic community as a program committee member for major data management conferences\, including VLDB\, SIGMOD\, and EuroSys.
URL:https://homecse.iitd.ac.in/event/next-generation-ai-enhanced-stream-processing/
CATEGORIES:Seminars
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Kolkata:20250514T100000
DTEND;TZID=Asia/Kolkata:20250514T110000
DTSTAMP:20260921T143430
CREATED:20250512T144512Z
LAST-MODIFIED:20250512T144600Z
UID:1589-1747216800-1747220400@homecse.iitd.ac.in
SUMMARY:Synthesis and Arithmetic of Quantum Circuits
DESCRIPTION:Speaker: Amolak Kalra (https://sites.google.com/view/amolakratankalra/home) \nAbstract: Efficient decomposition of a unitary operator U using words from a universal gate set G is a fundamental problem in quantum computing. The process by which this is achieved is called circuit synthesis. This problem arises naturally in the context of quantum circuit\ncompilation. In this talk\, I will introduce this problem and explain how one can use tools from number theory to solve it. I will then explain some recent results that build on this connection.
URL:https://homecse.iitd.ac.in/event/synthesis-and-arithmetic-of-quantum-circuits/
LOCATION:Bharti 501\, IIT Campus\, Hauz Khas\, New Delhi
CATEGORIES:Seminars
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Kolkata:20250508T140000
DTEND;TZID=Asia/Kolkata:20250508T150000
DTSTAMP:20260921T143430
CREATED:20250507T062717Z
LAST-MODIFIED:20250507T063432Z
UID:1581-1746712800-1746716400@homecse.iitd.ac.in
SUMMARY:Trading Prophets: How to trade multiple stocks optimally
DESCRIPTION:Speaker: Surbhi Rajput\, MSR Student\, CSE Dept.\, IIT Delhi \nAbstract:\n\nIn the (single stock) \emph{trading prophet} problem formulated by Correa et\nal.\ [2023]\, an online algorithm observes a sequence of prices of a stock.\nAt each step\, the algorithm can either buy the stock by paying the current\nprice if it doesn't already hold the stock\, or it can sell the currently\nheld stock and collect the current price as a reward. The goal of the\nalgorithm is to maximize its overall profit. Correa et al.\ showed that the\noptimal competitive ratio for this problem is $\nicefrac{1}{2}$ when the\nstock prices are identically and independently distributed.\nIn this talk\, I will discuss the simplifications and generalizations of\nCorrea et al.'s analysis\, which led us to generalize the model by allowing\nthe algorithm to trade multiple stocks. First\, we generalize the model to\n$(k\,\ell\, \ell')$-\textsc{Trading Prophet Problem}\, wherein there are $k$\nstocks in the market\, and the online algorithm can hold up to $\ell$ stocks\nat any time\, where $\ell \leq k$. The online algorithm competes against an\noffline algorithm that can hold at most $\ell' \leq \ell$ stocks at any\ntime. Under the assumption that prices of different stocks are independent\,\nwe show that\, for any $\ell$\, $\ell'$\, and $k$\, the optimal competitive\nratio of $(k\,\ell\, \ell')$-\textsc{Trading Prophet Problem} is\n$\min\left\{\frac{1}{2}\,\frac{\ell}{k}\right\}$.\nWe further generalize it to $\mathcal{M}$-\textsc{Trading Prophet Problem}\nover a matroid $\mathcal{M}$ on the set of $k$ stocks\, wherein the stock\nprices at any given time are possibly correlated (but are independent across\ntime). The algorithm is allowed to hold only a feasible subset of stocks at\nany time. We prove a tight bound of $\frac{1}{1+d}$ on the competitive ratio\nof the $\mathcal{M}$-\textsc{Trading Prophet Problem}\, where $d$ is the\n\textit{density} of the matroid.\nWe then consider the non-i.i.d.\ random order setting over a matroid\,\nwherein stock prices drawn independently from $n$ potentially different\ndistributions are presented in a uniformly random order. In this setting\, we\nachieve a competitive ratio of at least $\frac{1}{1+d} - \mathcal{O}\n\left(\frac{1}{n} \right)$\, where $d$ is the density of the matroid\,\nmatching the hardness result for i.i.d.\ instances as $n$ approaches\n$\infty$.\nOur analysis of the above problems is based on the following key insights.\nFirst\, any algorithm can be simulated by one that\, on each time step\, sells\n\emph{all} its currently held stocks before buying a suitable subset of\nstocks. Second\, we prove that the general problem reduces to a restriction\nwhere the expected price of every stock is zero.\nThird\, we reduce the problem in the random order non-i.i.d.\ setting to the\ni.i.d. setting by leveraging the fact that the outcome of sampling two\nobjects without replacement from a large set is almost identically\ndistributed as the outcome of sampling with replacement.
URL:https://homecse.iitd.ac.in/event/trading-prophets-how-to-trade-multiple-stocks-optimally/
LOCATION:Bharti 501\, IIT Campus\, Hauz Khas\, New Delhi
CATEGORIES:Seminars
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Kolkata:20250506T120000
DTEND;TZID=Asia/Kolkata:20250506T130000
DTSTAMP:20260921T143430
CREATED:20250503T125437Z
LAST-MODIFIED:20250503T125437Z
UID:1572-1746532800-1746536400@homecse.iitd.ac.in
SUMMARY:Power and limitations of quantum computation and quantum cryptography by Dr. Srijita Kundu
DESCRIPTION:Title: Power and limitations of quantum computation and quantum cryptography \nSpeaker: Dr. Srijita Kundu \nAbstract: Quantum computers are approaching practical viability\, and quantum cryptography is already being deployed for secure communication. Understanding the capabilities and limitations of these technologies is crucial for their effective use. \nMy research lies at the intersection of quantum complexity theory and cryptography. I focus on proving what quantum computation can and cannot do in concrete models such as query and communication complexity. In this talk\, I will share results in both directions:\n1. I will talk about direct product theorems for quantum communication complexity\, which are a useful lower bound technique for quantum communication protocols.\n2. I will talk about quantum proofs being more powerful than classical proofs in query complexity.\nAdditionally\, I will talk about quantum protocols for novel cryptographic tasks such as certified deletion and uncloneable encryption\, whose security can be proved using the communication direct product theorems. \nShort Bio: Srijita Kundu completed her PhD at the Centre for Quantum Technologies in the National University of Singapore in 2021\, under the supervision of Prof. Rahul Jain. Since 2022\, she has been a postdoctoral researcher at the Institute for Quantum Computing in the University of Waterloo. She is interested in quantum complexity theory and cryptography.
URL:https://homecse.iitd.ac.in/event/power-and-limitations-of-quantum-computation-and-quantum-cryptography-by-dr-srijita-kundu/
LOCATION:Bharti 501\, IIT Campus\, Hauz Khas\, New Delhi
CATEGORIES:Seminars
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Kolkata:20250501T110000
DTEND;TZID=Asia/Kolkata:20250501T120000
DTSTAMP:20260921T143430
CREATED:20250425T023119Z
LAST-MODIFIED:20250427T060150Z
UID:1521-1746097200-1746100800@homecse.iitd.ac.in
SUMMARY:Certifying Large Language Models with LLMCert
DESCRIPTION:Speaker: Isha Chaudhary \nAbstract: Large Language Models (LLMs) are increasingly deployed in critical systems\, e.g.\, healthcare and finance and can produce incorrect and biased responses. These can cause huge social and economic losses to the deploying agencies and their clients. Conventional studies are\, however\, insufficient to thoroughly evaluate LLMs\, as they cannot scale to a large number of possible inputs and provide no formal guarantees. Therefore\, we develop and present the first family of LLM certification frameworks\, LLMCert\, consisting of certifiers providing formal probabilistic guarantees for desirable properties such as correct LLM reasoning and fairness on prohibitively large distributions of prompts. Our certificates are quantitative — they consist of provably high-confidence\, tight bounds on the probability of desirable LLM responses for random prompts sampled from a distribution. We design and certify novel specifications for bias and knowledge comprehension in individual certifiers – LLMCert-B (https://certifyllm.com/) and LLMCert-C (https://arxiv.org/abs/2402.15929)\, respectively. We illustrate bias certification for distributions of prompts created by applying varying prefixes drawn from a prefix distribution to a given set of prompts. We consider prefix distributions for random token sequences\, mixtures of manual jailbreaks\, and jailbreaks in the LLM’s embedding space to certify bias. We obtain non-trivial certified bounds on the probability of unbiased responses of SOTA LLMs\, exposing their vulnerabilities over distributions of prompts generated from computationally inexpensive prefix distributions. \nFor knowledge comprehension certification\, we design and use novel distributions of knowledge comprehension prompts with natural noise\, using knowledge graphs. We certify SOTA LLMs over specifications arising in precision medicine and general question-answering. We show previously undiscovered vulnerabilities of SOTA LLMs owing to natural noise in prompts. We also establish the first performance hierarchies with formal guarantees among SOTA LLMs\, pertaining to question-answering in precision medicine. \n  \nBio: Isha Chaudhary is a third-year Ph.D. candidate at the Siebel School of Computing and Data Science\, University of Illinois Urbana-Champaign\, advised by Prof. Gagandeep Singh. Her research interest is broadly in trustworthy foundation models and neural networks for computer systems. She graduated from a B.Tech. in Electrical Engineering from IIT Delhi in 2022. For details about her work\, please check out: https://ishachaudhary.web.illinois.edu/.
URL:https://homecse.iitd.ac.in/event/certifying-large-language-models-with-llmcert/
LOCATION:Bharti 501\, IIT Campus\, Hauz Khas\, New Delhi
CATEGORIES:Seminars
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Kolkata:20250429T150000
DTEND;TZID=Asia/Kolkata:20250429T160000
DTSTAMP:20260921T143430
CREATED:20250425T023346Z
LAST-MODIFIED:20250427T060123Z
UID:1523-1745938800-1745942400@homecse.iitd.ac.in
SUMMARY:Computation-In-Memory based Edge-AI for Healthcare: A Cross-Layer Approach
DESCRIPTION:Speaker: Sumit Diware \nAbstract: Recent advancements in artificial intelligence (AI) have driven the emergence of real-world cognitive products and services\, which rely on neural networks to perform complex tasks. Edge computing for AI (edge-AI) combines data sources with local hardware that executes neural network computations\, to improve the response latency\, data privacy/security\, and service reliability. Computation-in-memory (CIM) offers an energy-efficient and compact alternative to conventional neural network hardware for edge-AI\, by enabling in-situ data processing with emerging memory technologies called memristors. Healthcare stands out as a key domain for CIM\, due to its critical impact on society and the need for energy-efficient\, compact hardware in healthcare edge applications. However\, developing AI models for healthcare that are effective\, accurate\, and can fully reap CIM benefits remains a significant challenge. Moreover\, memristors exhibit non-idealities that lead to errors during hardware execution. In this talk\, I will describe our cross-layer research approach and contributions towards addressing these challenges. We first create effective\, accurate\, and CIM-oriented AI models for two healthcare applications: electrocardiogram (ECG) classification and diabetic retinopathy screening. We then devise mitigation strategies against memristor non-idealities and develop a system-on-chip tapeout as a holistic solution that covers the entire abstraction layer stack from application to fabrication. \nShort Bio: Sumit Diware obtained Ph.D. from the Computer Engineering Group at Delft University of Technology (TU Delft)\, Netherlands\, and M.Tech. in VLSI Design Tools and Technology (VDTT) from IIT Delhi. His research focuses on artificial intelligence (AI) processing architectures\, with expertise in computation-in-memory\, neuromorphic computing\, emerging memory technologies\, hardware-algorithm co-design\, and system-on-chip (SoC) design/tapeout. He has authored/co-authored several publications in leading conferences such as DATE\, DAC\, and ICCAD\, as well as IEEE journals including TBioCAS and TETCI. For his doctoral work\, he recently received the European Design & Automation Association (EDAA) Outstanding Dissertation Award at DATE 2025 conference. Before his Ph.D.\, Sumit was a research assistant at the Karlsruhe Institute of Technology (KIT)\, Germany\, where he worked on multicore SoC architectures. Prior to that\, he worked at Qualcomm India as a part of IIT Delhi’s VDTT program\, focusing on SoC power management architecture.
URL:https://homecse.iitd.ac.in/event/computation-in-memory-based-edge-ai-for-healthcare-a-cross-layer-approach/
LOCATION:Bharti 501\, IIT Campus\, Hauz Khas\, New Delhi
CATEGORIES:Seminars
END:VEVENT
END:VCALENDAR