- talks - lists.cs.princeton.edu

Elizabeth Mieczkowski will present her General Exam "Quantifying Collaborative Trade-offs with Insights from Distributed Systems" on Tuesday, May 13, 2025 at 12:00 PM in FC 008.
by CS Grad Department 08 May '25

08 May '25

Elizabeth Mieczkowski will present her General Exam "Quantifying Collaborative Trade-offs with Insights from Distributed Systems" on Tuesday, May 13, 2025 at 12:00 PM in FC 008. Zoom Link: https://princeton.zoom.us/j/6756891992 Committee Members: Tom Griffiths (advisor), Benjamin Eysenbach, Wyatt Lloyd, Natalia Velez Abstract: Collaboration enables multi-agent systems to achieve goals beyond the capabilities of any single agent. However, coordination is inherently complex and dynamic. How can we identify the underlying principles that should guide the division of labor and resources in multi-agent teams, and formalize the trade-offs that emerge during collaboration? I argue that research from distributed computing systems provides a powerful interpretability framework for multi-agent collaboration. This literature focuses on building efficient, scalable, and fault-tolerant systems, while exposing fundamental constraints—such as communication overhead and consistency—when computation is shared across machines. I first extend Amdahl’s Law, which predicts how serial bottlenecks constrain parallel speedup, to naturalistic multi-agent tasks and confirm its validity in a set of human experiments. I then apply this model to large-scale MARL simulations in Overcooked, SMAC, and MPE, demonstrating that specialized policies emerge as a function of task parallelizability. Finally, I demonstrate how the model can benchmark existing algorithms and human collaborative strategies. Reading List: https://docs.google.com/document/d/12GKMv_n99TcUJvj8iua44eHGkUbQAz2hwX_zUYj… 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

[Machine Learning Seminar] May 9: Virtual Talk by Prof. Yitao Liang on Minecraft and Vision-Language-Action Models
by Emily C. Lawrence 07 May '25

07 May '25

This Friday, Prof. Yitao Liang from Peking University will give a virtual talk about his research on Vision-Language-Action Models and their applications to open-world games (e.g., Minecraft). Time : May 9, from 10:00 to 11:00 a.m. ET at [ https://princeton.zoom.us/j/97152277086 | https://princeton.zoom.us/j/97152277086 ] We will also stream the talk together at Friend Center 008 . Title: Instruction-Following Control in an Open World: Vision-Language-Action Models and Beyond Abstract: With the emergence of large language models, the debate about whether an instruction-following long-horizon control agent will appear has resurfaced. However, GPT's emerging abilities seem difficult to reproduce in domains that require sequential control. In this talk, we will introduce various efforts by our group and other well-known research labs to develop such a model in open-world game domains (e.g., Minecraft). Due to its extremely high degree of freedom, traditional multi-task data-driven methods are unsustainable (we cannot afford to train on thousands of tasks simultaneously, it is too expensive). One possible direction is to leverage LLMs to achieve flexible task planning. In this talk, we present latest research in this direction, and how to obtain a universal prompt-based generic policy learning. Due to models’s increasingly impressive data fitting capability, how to properly evaluate in open-ended domains has also become a new frontier. We will also discuss how generative evaluation could facilitate a more principled evaluation. Bio: Yitao Liang ( [ mailto:yitaol@pku.edu.cn | yitaol(a)pku.edu.cn ] ) is an assistant professor and Boya Fellow at Peking University. He obtained his Ph.D. degree in Computer Science from UCLA, advised by Prof. Guy Van den Broeck. His research interests span reasoning, machine learning and AI agents. His work has received recognition from top AI conferences; for example, the best-paper honorable mention from AAMAS 2016, the best paper from RL for Real Life workshop in ICML 2019, a best paper runner-up from the LLD workshop in NeurIPS 2017, a best paper from the TEACH workshop in ICML2023. He regularly serves as area chairs and senior area chairs in top venues e.g., NeurIPS, ICML. Best, Wenzhe Li & Chi Jin

1 0

Ruyu Yan will present her General Exam "Nonlinear Perspective Photography via Geometry-aware Scene Completion" on Friday, May 9, 2025 at 1:00 PM in CS 402.
by CS Grad Department 06 May '25

06 May '25

Ruyu Yan will present her General Exam "Nonlinear Perspective Photography via Geometry-aware Scene Completion" on Friday, May 9, 2025 at 1:00 PM in CS 402. Committee Members: Adam Finkelstein (advisor), Felix Heide, Szymon Rusinkiewicz Abstract: Designers of modern mobile cameras aim to capture high-quality images using linear perspective projection. However, this projection often fails to align with users' natural perception, which is influenced by nonlinear phenomena; it also does not support creative composition needs, such as depth compression along a long street or tilt-shift effects. We introduce a computational photography system that enables nonlinear perspective projection as a composition process during capture on mobile devices. Our system reconstructs a low-resolution 3D scene representation using signed distance field fusion, allowing users to explore perspective transformations with interactive preview in the camera viewfinder. During capture, the system guides users to capture additional data to overcome challenges such as occlusions and the limited field of view of the physical lens. We then employ a 2D Gaussian splatting method to reconstruct a geometry proxy of the scene, which we use for warping the captured photos and exporting high-fidelity photos with nonlinear perspective. Reading List: https://docs.google.com/document/d/1yaZAtQX7YuhOO9343ITD-ieHEApp8teCrU9aQCq… 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

David Braun will present his General Exam "EASY-AS-PIE: Evolutionary Algorithm SYnthesizer Architecture Search via Pre-trained Instrument Embeddings" on Thursday, May 8, 2025 at 10:00 AM in CS 402.
by gradinfo＠cs.princeton.edu 06 May '25

06 May '25

David Braun will present his General Exam "EASY-AS-PIE: Evolutionary Algorithm SYnthesizer Architecture Search via Pre-trained Instrument Embeddings" on Thursday, May 8, 2025 at 10:00 AM in CS 402. Committee Members: Adam Finkelstein (advisor), Tri Dao, Mae Milano Abstract: Software synthesizers offer immense creative possibilities for sound design, but navigation of their high-dimensional parameter spaces poses a challenge for musicians. This has motivated efforts toward automatic synthesizer programming-systems that generate parameter settings given a target sound or description. However, such systems require an automated means of comparing the similarity of sounds, which is complicated due to psychoacoustic factors. Existing methods often address synthesizer programming using gradient descent or genetic algorithms. Gradient-based optimization faces challenges because synthesizer signal processing is highly recurrent and can involve gradient discontinuities. Conversely, evolutionary methods handle such discontinuities but typically operate with fixed synthesizer architectures, limiting creative potential. Allowing dynamic architectural changes exponentially increases search complexity due to the combinatorial explosion in connecting synthesizer modules. Neither approach is well-suited for automatic programming of a large corpus of target sounds. Nonetheless, automatic synthesizer programming remains a valuable research direction because synthesizers offer greater interpretability compared to black-box neural audio synthesis models, enabling musicians and sound designers to intuitively understand and manipulate generated sounds. Motivated by these considerations, we propose a novel framework that uniquely combines two established concepts: evolutionary algorithms for synthesizer architecture search and pre-trained instrument embeddings (specifically CLAP) for perceptual comparison of sound similarity. Crucially, our method efficiently optimizes synthesizer presets for thousands of distinct target sounds simultaneously. Our evolutionary framework involves a hierarchical genetic representation comprising distinct gene classes for oscillator settings, envelopes, various audio effects, and more. Each parameter is modeled with dedicated numeric genes, either continuous (e.g., an oscillator's sustain level) or discrete (e.g., filter type), and mutations are performed by sampling from precomputed parameter distributions extracted from real synthesizer presets. Since the genome includes information about modulation routings (e.g., "Envelope 1 modulating Reverb Size with a Depth of 40%"), the genetic mutation and crossover operations explore the synthesizer architecture space in addition to a purely numeric parameter space. Our evolutionary framework also incorporates two complementary archive structures to balance exploration and exploitation. A grid-based archive inspired by MAP-Elites uses dimensionality reduction on CLAP embeddings, enabling structured diversity-driven exploration of the preset space, untethered to target sounds. Concurrently, a second archive maintains champion presets for each target sound, ranked using cosine distance metrics in the embedding space, facilitating exploitation. Through ablation studies, we demonstrate that this dual-archive approach effectively improves alignment with a held-out audio dataset measured by Kernel Audio Distance (KAD). Moreover, when optimizing for a single target sound, our approach can flexibly use a different loss signal such as a multi-scale spectrogram loss while still benefiting from pre-existing archives. We show that the addition of architecture search in such a scenario results in superior reconstruction of query sounds compared to only modifying numeric parameters. This research demonstrates that embedding-driven evolutionary strategies combined with flexible synthesizer representations enable automated and diverse exploration of sounds.

1 1

Haoran Wan will present his General Exam "Facilitating the Co-Design of Applications and Cellular Network both Internally and Externally" on May 12, 2025 at 12pm in CS 302.
by Gradinfo 05 May '25

05 May '25

Haoran Wan will present his General Exam "Facilitating the Co-Design of Applications and Cellular Network both Internally and Externally" on May 12, 2025 at 12pm in CS 302. Committee: Kyle Jamieson (adviser), Maria Apostolaki, Wyatt Lloyd Title: Facilitating the Co-Design of Applications and Cellular Network both Internally and Externally Abstract: 5G has emerged as the dominant cellular Radio Access Network (RAN) technology, serving as a critical last-mile solution for modern connectivity. Building on significant innovations in the physical (PHY) and medium access control (MAC) layers over LTE, 5G promises substantially higher throughput and lower latency. However, wireless networks suffer from highly variable end-to-end throughput due to user movement, channel fading and interference. Compounding this issue, the base station keeps deep packet buffers for retransmissions to mitigate packet loss caused by the varying channels, which paradoxically causes the high end-to-end latency. The variable throughput and hidden buffer pose performance challenges for many applications, ranging from congestion control for bulk downloading to interactive videos. And the closed nature of the RAN inhibits co-design of applications and the cellular network. To bridge the gap and facilitate the co-design of applications and the cellular network, we worked from two perspectives: 1) Externally, we developed a non-intrusive telemetry tool called NR-Scope to decode the control information passively from the raw radio signals, providing the MAC/PHY throughput of the base station in sub-millisecond level granularity; 2) Internally, we designed a base station module called L4Span that predicts the egress rate of the hidden queue and exposes this information through explicit congestion notification (ECN) bits in the IP header. Our work paves the way for co-designing future cellular systems and applications to fully leverage 5G’s capabilities.

1 0

Viola Chen will present her General Exam "Inferring Ancestral Cell States Using Transcriptome Geometry and Lineage Trees" on Thursday, May 8, 2025 at 10:30 AM in CS 302.
by CS Grad Department 05 May '25

05 May '25

Viola Chen will present her General Exam " Inferring Ancestral Cell States Using Transcriptome Geometry and Lineage Trees " on Thursday, May 8, 2025 at 10:30 AM in CS 302 . Committee Members: Benjamin Raphael (advisor), Mona Singh , Yuri Pritykin Abstract: Understanding how transcriptomic states evolve along a developmental lineage is a fundamental question in single-cell biology. While CRISPR-based lineage tracing technologies now enable high-resolution reconstruction of cell lineage trees, the transcriptomic states of internal (ancestral) nodes remain unobserved due to the destructive nature of sequencing. Existing methods for inferring these states either rely on strong parametric assumptions (e.g., Gaussian random walks in treeVAE) or discard true lineage information altogether (e.g., pseudotime methods). We present a geometry-aware parsimony framework for inferring ancestral transcriptomic states in cell lineage trees. Our approach leverages the principle of minimal transcriptomic change along a known tree structure, using either discrete or continuous representations of cell states. In the continuous case, we extend squared-change parsimony to non-Euclidean manifolds by computing Fréchet means under biologically meaningful distance metrics, such as graph geodesics derived from k-nearest neighbor (kNN) transcriptomic similarity graphs. We develop an efficient iterative optimization scheme, applicable to Euclidean, manifold, or graph-structured spaces, and benchmark our method on synthetic manifolds and real lineage-resolved datasets (e.g., C. elegans). Our results show improved recovery of internal cell states and differentiation trajectories compared to baseline methods. This framework provides a general and interpretable tool for reconstructing unobserved developmental states from lineage-traced single-cell data. Reading List: https://docs.google.com/document/d/1CtluDXuGFs2Z6IOfPlQg8gYGnNjemOPkjpYyuZZ… 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

Kart Kandula will present his General Exam "Competition by Design: Experimental Evidence on Search Engine Use and Choice Architecture Interventions" on May 6, 2025 at 2:30pm in 306 Sherrerd Hall.
by Gradinfo 02 May '25

02 May '25

Kart Kandula will present his General Exam "Competition by Design: Experimental Evidence on Search Engine Use and Choice Architecture Interventions" on May 6, 2025 at 2:30pm in 306 Sherrerd Hall. Committee: Jonathan Mayer (adviser), Peter Henderson, Andrés Monroy-Hernández All are welcome to attend the talk. Abstract: On both sides of the Atlantic, regulators have intensified scrutiny of Google’s dominance in the market for general search services, concerned that the company has engaged in exclusionary practices to maintain its position as the default gateway to the internet. In the United States, this culminated in a landmark ruling on August 5, 2024, in which a federal court found Google in violation of § 2 of the Sherman Act for unlawfully monopolizing the online search market. Now in the remedies phase, the U.S. Department of Justice has proposed a suite of interventions, including the use of “choice screens” to prompt users to select a default search engine. Yet critics argue that such measures, introduced too late and with limited enforcement mechanisms, fall short of meaningfully disrupting Google's entrenchment—an argument bolstered by the European Union’s tepid results following a similar Android-based remedy in 2019. To assess the real-world effectiveness of such interventions, this study aims to examine how default search engine settings shape user choices and how a variety of design interventions—modeled on plausible remedial frameworks available to regulators and courts —influence search behavior. We employ a two-phase experimental design targeting both desktop and mobile platforms, with participants recruited via Prolific. The first phase establishes a behavioral baseline while administering the intervention across the experimental conditions. The second phase, conducted after a delay, assesses the durability of observed behavioral shifts, enabling us to evaluate the long-term efficacy of choice architecture interventions. For mobile participants, we target iPhone users and implement choice architecture interventions using iOS Shortcuts, collecting behavioral data through Safari history exports. These methods, made possible by recent iOS updates, enable a novel and scalable approach to studying user behavior in naturalistic settings and open new possibilities for field experiments in digital platform research. Our findings will provide empirical insight into the real-world potency of behavioral remedies in digital platform markets—an area where policymaking has outpaced empirical evidence and measurement. Reading List 1. Dark Patterns and Privacy Choice 1. Jamie Luguri and Lior Jacob Strahilevitz. 2021. Shining a light on dark patterns. Journal of Legal Analysis 13, 1 (2021), 43–109. 2. Arunesh Mathur, Mihir Kshirsagar, and Jonathan Mayer. 2021. What makes a dark pattern . . . dark? Design attributes, normative considerations, and measurement methods. In Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems . ACM, New York, NY, USA, 1–18. 3. Christine Utz, Martin Degeling, Sascha Fahl, Florian Schaub, and Thorsten Holz. 2019. (Un)informed consent: Studying GDPR consent notices in the field. In Proceedings of the ACM SIGSAC Conference on Computer and Communications Security (CCS’19) . ACM, 973–990. 4. Midas Nouwens, Ilaria Liccardi, Michael Veale, David Karger, and Lalana Kagal. 2020. Dark patterns after the GDPR: Scraping consent pop-ups and demonstrating their influence. In Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems . 1–13. 2. Field Experiments Involving Online Activity 1. Hunt Allcott, Juan Camilo Castillo, Matthew Gentzkow, Leon Musolff, and Tobias Salz. 2024. Sources of Market Power in Web Search: Evidence from a Field Experiment. Working Paper (2024). 2. Andrew M Guess, Neil Malhotra, Jennifer Pan, Pablo Barberá, Hunt Allcott, Taylor Brown, Adriana Crespo-Tenorio, Drew Dimmery, Deen Freelon, Matthew Gentzkow, et al. 2023. How do social media feed algorithms affect attitudes and behavior in an election campaign? Science 381, 6656 (2023), 398–404. 3. Erik Brynjolfsson, Avinash Collis, and Felix Eggers. 2019. Using massive online choice experiments to measure changes in well-being. Proceedings of the National Academy of Sciences 116, 15 (2019), 7250–7255. 3. Human-Computer Interaction 1. Alma Whitten and J. Doug Tygar. 1999. Why Johnny can’t encrypt: A usability evaluation of PGP 5.0. In Proceedings of the USENIX Security Symposium . USENIX Association, Berkeley, CA, 1–16. Alan Dix, Janet Finlay, Gregory D. Abowd, and Russel Beale. 2004. Human-Computer Interaction (3rd ed.). Pearson Education Limited.

1 0

Kaya Alpturer will present his General Exam "Optimal RANDAO Manipulation" on May 12, 2025 at 12:30pm in CS 301.
by Gradinfo 02 May '25

02 May '25

Kaya Alpturer will present his General Exam "Optimal RANDAO Manipulation" on May 12, 2025 at 12:30pm in CS 301. Committee: Matt Weinberg (adviser), Mark Braverman, Pramod Viswanath All are welcome to attend the talk. Abstract: It is well-known that RANDAO manipulation is possible in Ethereum if an adversary controls the proposers assigned to the last slots in an epoch. We provide a methodology to compute, for any fraction α of stake owned by an adversary, the maximum fraction f(α) of rounds that a strategic adversary can propose. We further implement our methodology and compute f(⋅) for all α. For example, we conclude that an optimal strategic participant with 5% of the stake can propose a 5.048% fraction of rounds, 10% of the stake can propose a 10.19% fraction of rounds, and 20% of the stake can propose a 20.68% fraction of rounds. * Roughgarden, Tim. Twenty lectures on algorithmic game theory . Cambridge University Press, 2016. * Dwork, Cynthia, Nancy Lynch, and Larry Stockmeyer. "Consensus in the presence of partial synchrony." Journal of the ACM (JACM) 35.2 (1988): 288-323. * Fischer, Michael J., Nancy A. Lynch, and Michael S. Paterson. "Impossibility of distributed consensus with one faulty process." Journal of the ACM (JACM) 32.2 (1985): 374-382. * Yin, Maofan, et al. "HotStuff: BFT consensus with linearity and responsiveness." Proceedings of the 2019 ACM symposium on principles of distributed computing . 2019. * Cleve, Richard. "Limits on the security of coin flips when half the processors are faulty." Proceedings of the eighteenth annual ACM symposium on Theory of computing . 1986. * Eyal, Ittay, and Emin Gün Sirer. "Majority Is Not Enough: Bitcoin Mining Is Vulnerable." International Conference on Financial Cryptography and Data Security . Berlin, Heidelberg: Springer Berlin Heidelberg, 2014. * Sapirshtein, Ayelet, Yonatan Sompolinsky, and Aviv Zohar. "Optimal selfish mining strategies in bitcoin." Financial Cryptography and Data Security: 20th International Conference, FC 2016, Christ Church, Barbados, February 22–26, 2016, Revised Selected Papers 20 . Springer Berlin Heidelberg, 2017. * Cai, Yang, Nikhil R. Devanur, and S. Matthew Weinberg. "A duality based unified approach to bayesian mechanism design." Proceedings of the forty-eighth annual ACM symposium on Theory of Computing . 2016. * Beyhaghi, Hedyeh, and S. Matthew Weinberg. "Optimal (and benchmark-optimal) competition complexity for additive buyers over independent items." Proceedings of the 51st Annual ACM SIGACT Symposium on Theory of Computing . 2019. * Roughgarden, Tim. "Transaction Fee Mechanism Design." Journal of the ACM 71.4 (2024): 1-25. * Heilman, Ethan, et al. "Eclipse attacks on {Bitcoin’s}{peer-to-peer} network." 24th USENIX security symposium (USENIX security 15) . 2015. * Matheus V. X. Ferreira, S. Matthew Weinberg. “Proof-of-Stake Mining Games with Perfect Randomness” EC 2021.

1 0

Murali Srirangam Ramanujam will present his FPO "Leveraging Structural Stability for Efficient Compute-Memory Tradeoffs in Edge Systems" on 5/16/25 at 1:30pm in Friend 125.
by Gradinfo 02 May '25

02 May '25

Murali Srirangam Ramanujam will present his FPO "Leveraging Structural Stability for Efficient Compute-Memory Tradeoffs in Edge Systems" on 5/16/25 at 1:30pm in Friend 125. Examiners: Ravi Netravali (adviser), Maria Apostolaki, and Kai Li Readers: Wyatt Lloyd and Amit Levy Abstract: Edge systems - such as mobile devices, IoT nodes, and embedded sensors - increasingly operate in dynamic environments while facing severe resource constraints. Among these, compute is often the primary bottleneck, limited not only by hardware capacity but also by thermal and energy constraints inherent to untethered operation. A common strategy in systems design is to trade memory for compute, e.g., via caching or memoization. But on the edge, memory is also scarce, making indiscriminate reuse infeasible. This thesis is built on a central insight: despite the dynamic inputs and behaviors observed at runtime, most deployed applications - from system software to ML pipelines - rely on underlying structural stability. This stability arises from how these systems are developed: through modular frameworks, predictable control paths, and abstractions that preserve behavior across updates. In this dissertation, we show how to dissect and extract these stable computational elements and selectively reuse them to reduce compute overhead, even when memory is tightly limited. This allows us to affirmatively answer a fundamental question: can edge systems, without the luxury of abundant resources, still benefit from the kind of reuse strategies common in richer environments? First, we present Floo, a system that memoizes function-level computations in mobile applications by leveraging persistent interaction flows and function signatures, enabling effective reuse with minimal overheads. Then, we introduce Remembrall, which reuses parts of neural network pipelines in video analytics workloads by identifying redundant structure in learned embedding spaces - enabling lightweight adaptation and compute reduction under memory constraints. Together, these systems demonstrate that judicious use of limited memory, grounded in structural stability, can enable efficient compute reuse on the edge. All are welcome to attend.

1 0

Sunnie Kim will present herFPO "Advancing Responsible AI with Human-Centered Evaluation" on Wednesday, May 7, 2025 at 1:30 PM in CS 302 and Zoom.
by gradinfo＠cs.princeton.edu 02 May '25

02 May '25

Sunnie Kim will present herFPO "Advancing Responsible AI with Human-Centered Evaluation" on Wednesday, May 7, 2025 at 1:30 PM in CS 302 and Zoom. Location: Zoom link: https://princeton.zoom.us/my/sunniesuhyoung The members of Sunnies committee are as follows: Examiners: Olga Russakovsky (Adviser), Andrés Monroy-Hernández, Q. Vera Liao (Microsoft) Readers: Parastoo Abtahi, Jennifer Wortman Vaughan (Microsoft) A copy of her thesis is available upon request. Please email gradinfo(a)cs.princeton.edu if you would like a copy of the thesis. Everyone is invited to attend her talk. Abstract follows below: As Artificial Intelligence (AI) technologies are increasingly transforming how we live, work, and communicate, AI evaluation must take a human-centered approach to realistically reflect real-world performance and impact. In this thesis, I present research that advances human-centered evaluation, and subsequently responsible development and deployment of AI, by integrating knowledge and methods from AI and HumanComputer Interaction (HCI). First, using explainable AI as an example, I discuss the challenges and necessity of human (as opposed to automatic) evaluation and present human evaluation studies that identify new issues and opportunities in explainable AI research. Next, I bring context into the picture and present contextualized studies with real users that provide a holistic and nuanced understanding of users explainability needs and trust in AI. Then, I illustrate the importance of evaluating human-AI interaction through studies that examine how users trust, reliance, and other perceptions and behaviors are shaped by different features of AI technologies. In closing, I discuss the implications of these work, provide recommendations for human-centered evaluation, and describe future directions for responsible AI.

1 0