← Search

Christopher A. Choquette Choo

2 accepted papers

2026

Extracting alignment data in open models

ICML 2026poster

In this work, we show that it is possible to extract significant amounts of alignment training data from a post-trained model -- useful to steer the model to improve certain capabilities such as long-context reasoning, safety, instruction following, and maths. While the majority of related work on m…

Cited by 0SourceScholar
2022

The Fundamental Price of Secure Aggregation in Differentially Private Federated Learning

ICML 2022spotlight

We consider the problem of training a $d$ dimensional model with distributed differential privacy (DP) where secure aggregation (SecAgg) is used to ensure that the server only sees the noisy sum of $n$ model updates in every training round. Taking into account the constraints imposed by SecAgg, we c…