Jessica Sorrell

Postdoctoral Researcher
School of Engineering and Applied Science
University of Pennsylvania
jsorrell@seas.upenn.edu

I'm on the job market this year!

I'm a postdoc at UPenn, working with Aaron Roth and Michael Kearns. I completed my PhD at UCSD, advised by Daniele Micciancio and Russell Impagliazzo. I'm broadly interested in the theoretical foundations of responsible computing, particularly questions related to fairness, privacy, robustness, and reproducibility. I also work on problems in lattice-based cryptography, with a focus on secure computation.

Publications and Manuscripts

Replicable Reinforcement Learning

Eric Eaton, Marcel Hussing, Michael Kearns, Jessica Sorrell

To appear, NeurIPS 2023

[arXiv]

Stability is Stable: Connections between Replicability, Privacy, and Adaptive Generalization

Mark Bun, Marco Gaboardi, Max Hopkins, Russell Impagliazzo, Rex Lei, Toniann Pitassi, Satchit Sivakumar, Jessica Sorrell

STOC 2023

[arXiv]

Multicalibration as Boosting for Regression

Ira Globus-Harris, Declan Harrison, Michael Kearns, Aaron Roth, Jessica Sorrell

ICML 2023

[arXiv]

Securing Approximate Homomorphic Encryption Using Differential Privacy

Baiyu Li, Daniele Micciancio, Mark Schultz, Jessica Sorrell

Crypto 2022

[ePrint]

Reproducibility in Learning

Russell Impagliazzo, Rex Lei, Toniann Pitassi, Jessica Sorrell

STOC 2022

[arXiv]

Boosting in the Presence of Massart Noise

Ilias Diakonikolas, Russell Impagliazzo, Daniel Kane, Rex Lei, Jessica Sorrell, Christos Tzamos

COLT 2021

[arXiv]

Simpler Statistically Sender Private Oblivious Transfer from Ideals of Cyclotomic Integers

Daniele Micciancio, Jessica Sorrell

Asiacrypt 2020

[eprint]

Efficient, Noise-Tolerant, and Private Learning via Boosting

Mark Bun, Marco Leandro Carmosino, Jessica Sorrell

COLT 2020

[arXiv]

The Fiat-Shamir Zoo: Relating the Security of Different Signature Variants

Matilda Backendal, Mihir Bellare, Jessica Sorrell, Jiahao Sun

NordSec 2018

[eprint]

Ring Packing and Amortized FHEW Bootstrapping

Daniele Micciancio, Jessica Sorrell

ICALP 2018

[eprint]

Jessica Sorrell

Publications and Manuscripts

Replicable Reinforcement Learning

Stability is Stable: Connections between Replicability, Privacy, and Adaptive Generalization

Multicalibration as Boosting for Regression

Securing Approximate Homomorphic Encryption Using Differential Privacy

Reproducibility in Learning

Boosting in the Presence of Massart Noise

Simpler Statistically Sender Private Oblivious Transfer from Ideals of Cyclotomic Integers

Efficient, Noise-Tolerant, and Private Learning via Boosting

The Fiat-Shamir Zoo: Relating the Security of Different Signature Variants

Ring Packing and Amortized FHEW Bootstrapping

Interests