AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Uncertainty Calibration for Ensemble-Based Debiasing Methods

Ruibin Xiong, Yimeng Chen, Liang Pang, Xueqi Cheng, Zhi-Ming Ma, Yanyan Lan

NeurIPS 2021 · Label-efficient and robust learning › Recognition and long-tail · approach: Confidence calibration objectives · 3 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
Add Dirichlet calibration of the bias-only modelPoE · FEVER Symmetric v1accuracy: 62.0 -> 65.9Improved
Add Dirichlet calibration of the bias-only modelPoE · FEVER Symmetric v2accuracy: 65.9 -> 69.1Improved
Add Dirichlet calibration of the bias-only modelLearned-Mixin · FEVER Symmetric v1accuracy: 59.8 -> 61.5Improved
Add Dirichlet calibration of the bias-only modelPoE · FEVERin-distribution accuracy: 84.0 -> 87.1Improved
Add temperature scaling calibration of the bias-only modelPoE · FEVER Symmetric v1accuracy: 62.0 -> 63.3Improved
Use Dirichlet instead of temperature scaling calibratorPoE · FEVER Symmetric v1accuracy: 63.3 -> 65.9Improved
Add Dirichlet calibration of the bias-only modelPoE · MNLI syntactic bias, HANSaccuracy: 68.1 -> 70.7Improved
Add Dirichlet calibration of the bias-only modelLearned-Mixin · MNLI syntactic bias, HANSaccuracy: 65.5 -> 62.7Hurt
Add Dirichlet calibration of the bias-only modelLearned-Mixin · MNLI hypothesis-only, Hard-SPaccuracy: 78.2 -> 83.2Improved
Add Dirichlet calibration of the bias-only modelPoE · MNLI hypothesis-only, Hard-CDaccuracy: 79.4 -> 79.4No clear effect
Add Dirichlet calibration of the bias-only modelPoE · MNLI hypothesis-only, in-distributionaccuracy: 83.2 -> 82.7Hurt
Apply Dirichlet calibration to bias-only modelbias-only model · FEVERclasswise ECE: 7.11 -> 1.73Improved

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