Privacy-preserving machine learning aims to train models on private data...
In order to train networks for verified adversarial robustness, previous...
The ability to generate privacy-preserving synthetic versions of sensiti...
Recent works have tried to increase the verifiability of adversarially
t...
We propose a general framework for verifying input-output specifications...
Reliable detection of out-of-distribution (OOD) inputs is increasingly
u...
Recent improvements in large-scale language models have driven progress ...
Neural networks are part of many contemporary NLP systems, yet their
emp...
Adversarial training is an effective methodology for training deep neura...
Recent work has uncovered the interesting (and somewhat surprising) find...
Prior work on neural network verification has focused on specifications ...
While deep learning has led to remarkable results on a number of challen...
Recent works have shown that it is possible to train models that are
ver...
This paper proposes a new algorithmic framework,predictor-verifier
train...
This paper addresses the problem of formally verifying desirable propert...