AI Paper Atlas

Research map / Learning theory and algorithms / Fairness

SenSeI: Sensitive Set Invariance for Enforcing Individual Fairness

Mikhail Yurochkin, Yuekai Sun

ICLR 2021 · Learning theory and algorithms › Fairness · approach: Fairness-constrained classification · 18 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Use SenSeI fair regularization instead of ERMBERT-based toxic comment classifier · Toxic comment detectionCTF score: 0.072±0.002 -> 0.029±0.005Improved
Use SenSeI fair regularization instead of ERMBERT-based toxic comment classifier · Toxic comment detectionbalanced accuracy: 0.807±0.002 -> 0.791±0.005Hurt
Use SenSeI fair regularization instead of CLPBERT-based toxic comment classifier · Toxic comment detectionprediction consistency: 0.763±0.048 -> 0.773±0.043Improved
Use SenSeI fair regularization instead of SenSRBERT-based toxic comment classifier · Toxic comment detectionprediction consistency: 0.729±0.044 -> 0.773±0.043Improved
Use SenSeI fair regularization instead of ERMBERT-based bios occupation classifier · Bios occupation predictionbalanced accuracy: 0.842±0.002 -> 0.843±0.003Improved
Use SenSeI fair regularization instead of ERMBERT-based bios occupation classifier · Bios occupation predictionCTF score: 0.028±0.001 -> 0.003±0.000Improved
Use SenSeI fair regularization instead of ERMAdult income classifier · AdultGap_G RMS: .179 -> .043Improved
Use SenSeI fair regularization instead of ERMAdult income classifier · Adultbalanced accuracy (%): 82.9 -> 76.8Hurt
Increase fair regularization strength ρSenSeI occupation classifier · Bios occupation predictionprediction consistencyImproved
Use worst-case perturbation search rather than random counterfactual samplingSenSeI instead of CLP · Toxic comment detectionprediction consistency: 0.763±0.048 -> 0.773±0.043Improved

Later papers that compared with this method

2 later accepted papers report SenSeI as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy