no code implementations • 9 Jan 2024 • Gal Yona, Roee Aharoni, Mor Geva
In this work, we propose GRANOLA QA, a novel evaluation setting where a predicted answer is evaluated in terms of accuracy and informativeness against a set of multi-granularity answers.
no code implementations • 12 May 2023 • Gal Yona, Or Honovich, Itay Laish, Roee Aharoni
We use CID to highlight context-specific biases that are hard to detect with standard decoding strategies and quantify the effect of different input perturbations.
no code implementations • 28 Nov 2022 • Yoav Wald, Gal Yona, Uri Shalit, Yair Carmon
This suggests that the phenomenon of ``benign overfitting," in which models generalize well despite interpolating, might not favorably extend to settings in which robustness or fairness are desirable.
no code implementations • 25 Oct 2022 • Gal Yona, Amir Feder, Itay Laish
An important component in deploying machine learning (ML) in safety-critic applications is having a reliable measure of confidence in the ML model's predictions.
no code implementations • 10 Apr 2022 • Gal Yona, Shay Moran, Gal Elidan, Amir Globerson
We show that there is a natural class where this approach is sub-optimal, and that there is a more comparison-efficient active learning scheme.
no code implementations • 18 Mar 2022 • Guy N. Rothblum, Gal Yona
We formalize a natural (distribution-free) solution concept: given anticipated miscalibration of $\alpha$, we propose using the threshold $j$ that minimizes the worst-case regret over all $\alpha$-miscalibrated predictors, where the regret is the difference in clinical utility between using the threshold in question and using the optimal threshold in hindsight.
no code implementations • 27 Oct 2021 • Gal Yona, Daniel Greenfeld
They argue that some popular saliency methods should not be used for explainability purposes since the maps they produce are not sensitive to the underlying model that is to be explained.
no code implementations • 2 Oct 2021 • Guy N. Rothblum, Gal Yona
The notion of "too much" is quantified via a parameter $\gamma$ that serves as a vehicle for specifying acceptable tradeoffs between accuracy and fairness, in a way that is independent from the specific metrics used to quantify fairness and accuracy in a given task.
no code implementations • 20 May 2021 • Guy N Rothblum, Gal Yona
An agnostic PAC learning algorithm finds a predictor that is competitive with the best predictor in a benchmark hypothesis class, where competitiveness is measured with respect to a given loss function.
no code implementations • 26 Nov 2020 • Cynthia Dwork, Michael P. Kim, Omer Reingold, Guy N. Rothblum, Gal Yona
Prediction algorithms assign numbers to individuals that are popularly understood as individual "probabilities" -- what is the probability of 5-year survival after cancer diagnosis?
no code implementations • 9 Oct 2019 • Gal Yona, Amirata Ghorbani, James Zou
We propose Extended Shapley as a principled framework for this problem, and experiment empirically with how it can be used to address questions of ML accountability.
no code implementations • 3 Apr 2019 • Michael P. Kim, Aleksandra Korolova, Guy N. Rothblum, Gal Yona
We introduce and study a new notion of preference-informed individual fairness (PIIF) that is a relaxation of both individual fairness and envy-freeness.
no code implementations • ICML 2018 • Guy N. Rothblum, Gal Yona
We show that approximate metric-fairness {\em does} generalize, and leverage these generalization guarantees to construct polynomial-time PACF learning algorithms for the classes of linear and logistic predictors.