Aaron Horowitz

Artificial Intelligence · 2021–2022

A researcher in the Science corpus: 2 papers; trusts 6 researchers and is trusted by 0 researchers here, as inferred from the citation graph.

Subfields

• Artificial Intelligence — 45 citations into it
• Safety, Robustness & Interpretability — 35 citations into it
• Evaluation & Benchmarks — 13 citations into it
• Computer Vision and Pattern Recognition — 7 citations into it
• Information Systems — 5 citations into it

Whom Aaron Horowitz trusts overall

Their strongest outbound trust, strongest first, inferred from their citations.

1. Allyson Ettinger
2. Thomas Liao
3. Emily M. Bender
4. Milad Moradi

Papers from Aaron Horowitz's perspective

Ordered by how strongly the researchers Aaron Horowitz trusts endorse each paper. Same corpus, a different order for every researcher.

1. Model Cards for Model Reporting
2. Gender Shades: Intersectional Accuracy Disparities in Commercial Gender Classification
3. Explaining and Harnessing Adversarial Examples
4. Fair Prediction with Disparate Impact: A Study of Bias in Recidivism Prediction Instruments
5. Datasheets for datasets
6. Certifying and Removing Disparate Impact
7. Inherent Trade-Offs in the Fair Determination of Risk Scores
8. To predict and serve?
9. Release Strategies and the Social Impacts of Language Models
10. Pervasive Label Errors in Test Sets Destabilize Machine Learning Benchmarks (NeurIPS Datasets and Benchmarks)

Their papers

• It's COMPASlicated: The Messy Relationship between RAI Datasets and Algorithmic Fairness Benchmarks (2021) — 118 citations
• The Fallacy of AI Functionality (2022) — 292 citations