no code implementations • 30 Apr 2024 • Nicholas S. Kersting, Mohammad Rahman, Suchismitha Vedala, Yang Wang
We introduce an intuitive method to test the robustness (stability and explainability) of any black-box LLM in real-time, based upon the local deviation from harmoniticity, denoted as $\gamma$.
no code implementations • 29 Apr 2024 • Nicholas S. Kersting, Yi Li, Aman Mohanty, Oyindamola Obisesan, Raphael Okochu
We introduce Harmonic Robustness, a powerful and intuitive method to test the robustness of any machine-learning model either during training or in black-box real-time inference monitoring without ground-truth labels.