galyona

About

Hi! My name is Gal Yona. I am currently a Research Scientist at Google Research Tel Aviv, where I work on making large language models more robust, factual and trustworthy. Before Google, I completed a PhD in Computer Science at the Weizmann Institute of Science, under the supervision of Guy Rothblum, where I focused on formalizing fairness & non-discrimination in machine learning algorithms.

Contact

You can contact me at: gal.yona at gmail.com

Publications

For an up to date list, check out my Google Scholar.

Confidence Improves Self-Consistency in LLMs‏‏
Amir Taubenfeld, Tom Sheffer, Eran Ofek, Amir Feder, Ariel Goldstein, Zorik Gekhman and Gal Yona
preprint [paper]

Keep Guessing? When Considering Inference Scaling, Mind the Baselines‏‏
Gal Yona, Or Honovich, Omer Levy and Roee Aharoni‏
Nations of the Americas Chapter of the Association for Computational Linguistics (NAACL) 2025 Findings [paper]

Can Large Language Models Faithfully Express Their Intrinsic Uncertainty in Words?‏
Gal Yona, Roee Aharoni and Mor Geva
Conference on Empirical Methods in Natural Language Processing (EMNLP) 2024 [paper]

Does Fine-Tuning LLMs on New Knowledge Encourage Hallucinations?‏‏
Zorik Gekhman, Gal Yona, Roee Aharoni, Matan Eyal, Amir Feder, Roi Reichart and Jonathan Herzig
Conference on Empirical Methods in Natural Language Processing (EMNLP) 2024 [paper]

Narrowing the Knowledge Evaluation Gap: Open-Domain Question Answering with Multi-Granularity Answers‏
Gal Yona, Roee Aharoni and Mor Geva
Association for Computational Linguistics (ACL) 2024 [paper]

Surfacing Biases in Large Language Models using Contrastive Input Decoding‏
Gal Yona, Or Honovich, Itay Laish and Roee Aharoni‏
preprint [paper]

Malign Overfitting: Interpolation Can Provably Preclude Invariance
Yoav Wald, Gal Yona, Uri Shalit and Yair Carmon
International Conference on Learning Representations (ICLR) 2023 [paper]

Decision-Making under Miscalibration
Guy N. Rothblum and Gal Yona
Innovations in Theoretical Computer Science (ITCS) 2023 [paper] [video] [code]

Useful Confidence Measures: Beyond the Max Score
Gal Yona, Amir Feder and Itay Laish
Workshop on Distribution Shifts at Neurips 2022 [paper]

Active Learning with Label Comparisons
Gal Yona, Shay Moran, Gal Elidan and Amir Globerson
Uncertainty in Artificial Intelligence (UAI) 2022 [paper] [poster]

Beyond Bernoulli: Generating Random Outcomes that cannot be Distinguished from Nature
Cynthia Dwork, Michael P. Kim, Omer Reingold, Guy N. Rothblum and Gal Yona
Algorithmic Learning Theory (ALT) 2022 [paper]

On Fairness and Stability in Two-Sided Matchings
Gili Karni, Guy N. Rothblum and Gal Yona
Innovations in Theoretical Computer Science (ITCS) 2022 [paper]

Revisiting Sanity Checks for Saliency Maps
Gal Yona and Daniel Greenfeld
Workshop on eXplainable AI approaches for debugging and diagnosis at Neurips 2021 [paper] [poster]

Consider the Alternatives: Navigating Fairness-Accuracy Tradeoffs via Disqualification
Guy N. Rothblum and Gal Yona
ACM Conference on Equity and Access in Algorithms, Mechanisms, and Optimization (EAAMO) 2021 (Poster) [paper] [poster]

Multi-group Agnostic PAC Learning
Guy N. Rothblum and Gal Yona
International Conference on Machine Learning (ICML) 2021 [paper] [video]

Who’s responsible? Jointly quantifying the contribution of the learning algorithm and training data
Gal Yona, Amirata Ghorbani and James Zou
Artificial Intelligence, Ethics and Society (AIES) 2021 [paper] [poster]

Outcome Indistinguishability
Cynthia Dwork, Michael P. Kim, Omer Reingold, Guy N. Rothblum and Gal Yona \
Symposium on Theory of Computing (STOC) 2021 [paper] [video]

Addressing bias in prediction models by improving subpopulation calibration
Noam Barda, Gal Yona, Guy N Rothblum, Philip Greenland, Morton Leibowitz, Ran Balicer, Eitan Bachmat and Noa Dagan Journal of the American Medical Informatics Association (JAMIA) 2020 [paper]

Developing a COVID-19 mortality risk prediction model when individual-level data are not available
Noam Barda, Dan Riesel, Amichay Akriv, Joseph Levy, Uriah Finkel, Gal Yona, Daniel Greenfeld, Shimon Sheiba, Jonathan Somer, Eitan Bachmat, Guy N Rothblum, Uri Shalit, Doron Netzer, Ran Balicer and Noa Dagan
Nature Communications 2020 [paper]

Preference-Informed Fairness
Michael P. Kim, Aleksandra Korolova, Guy N. Rothblum and Gal Yona
Innovations in Theoretical Computer Science (ITCS) 2020 [paper] [talk]

Evidence-Based Rankings
Cynthia Dwork, Michael P. Kim, Omer Reingold, Guy N. Rothblum and Gal Yona
Foundations of Computer Science (FOCS) 2019 [paper]

Probably Approximately Metric Fair Learning
Guy N. Rothblum and Gal Yona
International Conference on Machine Learning (ICML) 2018
[paper] [talk]