Abhishek Shetty
YOU?
Author Swipe
View article: Dimension-Free Correlated Sampling for the Hypersimplex
Dimension-Free Correlated Sampling for the Hypersimplex Open
Sampling from multiple distributions so as to maximize overlap has been studied by statisticians since the 1950s. Since the 2000s, such correlated sampling from the probability simplex has been a powerful building block in disparate areas …
View article: Taming Imperfect Process Verifiers: A Sampling Perspective on Backtracking
Taming Imperfect Process Verifiers: A Sampling Perspective on Backtracking Open
Test-time algorithms that combine the generative power of language models with process verifiers that assess the quality of partial generations offer a promising lever for eliciting new reasoning capabilities, but the algorithmic design sp…
View article: Physics of unravelling and micromechanics of hagfish threads
Physics of unravelling and micromechanics of hagfish threads Open
Hagfish slime is a unique biological material composed of mucus and protein threads that rapidly deploy into a cohesive network when deployed in seawater. The forces involved in thread deployment and interactions among mucus and threads ar…
View article: The Space Complexity of Learning-Unlearning Algorithms
The Space Complexity of Learning-Unlearning Algorithms Open
We study the memory complexity of machine unlearning algorithms that provide strong data deletion guarantees to the users. Formally, consider an algorithm for a particular learning task that initially receives a training dataset. Then, aft…
View article: Beyond Worst-Case Online Classification: VC-Based Regret Bounds for Relaxed Benchmarks
Beyond Worst-Case Online Classification: VC-Based Regret Bounds for Relaxed Benchmarks Open
We revisit online binary classification by shifting the focus from competing with the best-in-class binary loss to competing against relaxed benchmarks that capture smoothed notions of optimality. Instead of measuring regret relative to th…
View article: Pilot‐Scale Production of Spray Dried Pulse Protein‐Polyphenol Particles and Protein Bars
Pilot‐Scale Production of Spray Dried Pulse Protein‐Polyphenol Particles and Protein Bars Open
Pulses are still under‐consumed in the USA and in most of the Western world. Scalable, industrially friendly production strategies that deliver good‐tasting pulse ingredients and products are necessary to intensify the use of pulse protein…
View article: Small Loss Bounds for Online Learning Separated Function Classes: A Gaussian Process Perspective
Small Loss Bounds for Online Learning Separated Function Classes: A Gaussian Process Perspective Open
In order to develop practical and efficient algorithms while circumventing overly pessimistic computational lower bounds, recent work has been interested in developing oracle-efficient algorithms in a variety of learning settings. Two such…
View article: 3D-Printable Elastomers for Real-Time Autonomous Self-Healing in Soft Devices
3D-Printable Elastomers for Real-Time Autonomous Self-Healing in Soft Devices Open
Photocurable self-healing elastomers are promising candidates for producing complex soft devices that can mend damage. However, the practicality of these materials is limited by reliance on external stimuli, custom synthesis, manual realig…
View article: Stress-controlled medium-amplitude oscillatory shear (MAOStress) of PVA–Borax
Stress-controlled medium-amplitude oscillatory shear (MAOStress) of PVA–Borax Open
We report the first-ever complete measurement of MAOStress material functions, which reveal that stress can be more fundamental than strain or strain rate for understanding linearity limits as a function of Deborah number. The material use…
View article: Biopolymer networks packed with microgels combine strain stiffening and shape programmability
Biopolymer networks packed with microgels combine strain stiffening and shape programmability Open
Biomaterials that can be reversibly stiffened and shaped could be useful in broad biomedical applications where form-fitting scaffolds are needed. Here we investigate the combination of strong non-linear elasticity in biopolymer networks w…
View article: Tolerant Algorithms for Learning with Arbitrary Covariate Shift
Tolerant Algorithms for Learning with Arbitrary Covariate Shift Open
We study the problem of learning under arbitrary distribution shift, where the learner is trained on a labeled set from one distribution but evaluated on a different, potentially adversarially generated test distribution. We focus on two f…
View article: Identifying rheological regimes within pyroclastic density currents
Identifying rheological regimes within pyroclastic density currents Open
Pyroclastic density currents (PDCs) are the most lethal of all volcanic hazards. An ongoing challenge is to accurately forecast their run-out distance such that effective mitigation strategies can be implemented. Central to this goal is an…
View article: Jamming Memory into Acoustically Trained Dense Suspensions under Shear
Jamming Memory into Acoustically Trained Dense Suspensions under Shear Open
Systems driven far from equilibrium often retain structural memories of their processing history. This memory has, in some cases, been shown to dramatically alter the material response. For example, work hardening in crystalline metals can…
View article: Jamming memory into acoustically trained dense suspensions under shear
Jamming memory into acoustically trained dense suspensions under shear Open
Systems driven far from equilibrium often retain structural memories of their processing history. This memory has, in some cases, been shown to dramatically alter the material response. For example, work hardening in crystalline metals can…
View article: On the Performance of Empirical Risk Minimization with Smoothed Data
On the Performance of Empirical Risk Minimization with Smoothed Data Open
In order to circumvent statistical and computational hardness results in sequential decision-making, recent work has considered smoothed online learning, where the distribution of data at each time is assumed to have bounded likeliehood ra…
View article: Oracle-Efficient Differentially Private Learning with Public Data
Oracle-Efficient Differentially Private Learning with Public Data Open
Due to statistical lower bounds on the learnability of many function classes under privacy constraints, there has been recent interest in leveraging public data to improve the performance of private learning algorithms. In this model, algo…
View article: Characterization of the temperature and frequency dependency of the complex Poisson’s ratio using a novel combined torsional-axial rheometer
Characterization of the temperature and frequency dependency of the complex Poisson’s ratio using a novel combined torsional-axial rheometer Open
This study discusses the feasibility of using a combined torsional-axial rheometer to indirectly measure the complex Poisson’s ratio based on shear and Young’s modulus. For this purpose, isothermal frequency sweeps in torsion and extension…
View article: Smooth Nash Equilibria: Algorithms and Complexity
Smooth Nash Equilibria: Algorithms and Complexity Open
A fundamental shortcoming of the concept of Nash equilibrium is its computational intractability: approximating Nash equilibria in normal-form games is PPAD-hard. In this paper, inspired by the ideas of smoothed analysis, we introduce a re…
View article: The manifold rheology of fluidized granular media
The manifold rheology of fluidized granular media Open
Fluidized granular media have a rich rheology: measuring shear stress $σ$ as a function of shear rate $\dotγ$, they exhibit Newtonian behavior $σ\sim\dotγ$ for low densities and shear rates, develop a yield stress for intermediate shear ra…
View article: Adversarial Resilience in Sequential Prediction via Abstention
Adversarial Resilience in Sequential Prediction via Abstention Open
We study the problem of sequential prediction in the stochastic setting with an adversary that is allowed to inject clean-label adversarial (or out-of-distribution) examples. Algorithms designed to handle purely stochastic data tend to fai…
View article: Optimal PAC Bounds Without Uniform Convergence
Optimal PAC Bounds Without Uniform Convergence Open
In statistical learning theory, determining the sample complexity of realizable binary classification for VC classes was a long-standing open problem. The results of Simon and Hanneke established sharp upper bounds in this setting. However…
View article: Smoothed Analysis of Sequential Probability Assignment
Smoothed Analysis of Sequential Probability Assignment Open
We initiate the study of smoothed analysis for the sequential probability assignment problem with contexts. We study information-theoretically optimal minmax rates as well as a framework for algorithmic reduction involving the maximum like…
View article: Progressive Ensemble Distillation: Building Ensembles for Efficient Inference
Progressive Ensemble Distillation: Building Ensembles for Efficient Inference Open
We study the problem of progressive ensemble distillation: Given a large, pretrained teacher model $g$, we seek to decompose the model into smaller, low-inference cost student models $f_i$, such that progressively evaluating additional mod…
View article: Understanding the transient large amplitude oscillatory shear behavior of yield stress fluids
Understanding the transient large amplitude oscillatory shear behavior of yield stress fluids Open
A full understanding of the sequence of processes exhibited by yield stress fluids under large amplitude oscillatory shearing is developed using multiple experimental and analytical approaches. A novel component rate Lissajous curve, where…
View article: The One-Inclusion Graph Algorithm is not Always Optimal
The One-Inclusion Graph Algorithm is not Always Optimal Open
The one-inclusion graph algorithm of Haussler, Littlestone, and Warmuth achieves an optimal in-expectation risk bound in the standard PAC classification setup. In one of the first COLT open problems, Warmuth conjectured that this predictio…
View article: A comparative study between USG guided aspiration and incision and drainage of breast abscess
A comparative study between USG guided aspiration and incision and drainage of breast abscess Open
Introduction: Breast abscess is a major cause of morbidity in developing countries especially among lactating mothers. It is the most common benign breast problem in pregnancy and puerperium. The aim of this study is to evaluate the effica…
View article: Oracle-Efficient Online Learning for Beyond Worst-Case Adversaries
Oracle-Efficient Online Learning for Beyond Worst-Case Adversaries Open
In this paper, we study oracle-efficient algorithms for beyond worst-case analysis of online learning. We focus on two settings. First, the smoothed analysis setting of [RST11,HRS22] where an adversary is constrained to generating samples …
View article: Influence of particle size on powder rheology and effects on mass flow during directed energy deposition additive manufacturing
Influence of particle size on powder rheology and effects on mass flow during directed energy deposition additive manufacturing Open
View article: Matrix Discrepancy from Quantum Communication
Matrix Discrepancy from Quantum Communication Open
We develop a novel connection between discrepancy minimization and (quantum) communication complexity. As an application, we resolve a substantial special case of the Matrix Spencer conjecture. In particular, we show that for every collect…
View article: Universal scaling of shear thickening transitions
Universal scaling of shear thickening transitions Open
Nearly all dense suspensions undergo dramatic and abrupt thickening transitions in their flow behaviour when sheared at high stresses. Such transitions occur when the dominant interactions between the suspended particles shift from hydrody…