AI Paper Atlas

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

Learning to Predict Trustworthiness with Steep Slope Loss

Yan Luo, Yongkang Wong, Mohan Kankanhalli, Qi Zhao

NeurIPS 2021 · Label-efficient and robust learning › Recognition and long-tail · approach: Group-robust classification · 4 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
Replace TCP confidence loss with steep slope lossTrustworthiness oracle with ViT backbone and ViT classifier · ImageNet valAUC: 60.14±0.47 -> 73.68±0.27Improved
Replace TCP confidence loss with steep slope lossTrustworthiness oracle with ViT backbone and ResNet classifier · ImageNet valAUC: 66.23±1.02 -> 74.31±0.21Improved
Replace TCP confidence loss with steep slope lossTrustworthiness oracle with ResNet backbone and ViT classifier · ImageNet valAUC: 55.55±0.46 -> 64.34±0.16Improved
Replace TCP confidence loss with steep slope lossTrustworthiness oracle with ResNet backbone and ResNet classifier · ImageNet valAUC: 62.73±0.14 -> 67.92±0.11Improved
Replace TCP confidence loss with steep slope lossViT trustworthiness oracle · Stylized ImageNet valAUC: 57.29 -> 67.68Improved
Replace TCP confidence loss with steep slope lossViT trustworthiness oracle · Adversarial ImageNet valAUC: 55.81 -> 65.36Improved
Replace TCP confidence loss with steep slope lossViT trustworthiness oracle · ImageNet-C Defocus blurAUC: 54.27 -> 64.34Improved
Replace TCP confidence loss with steep slope lossViT trustworthiness oracle · ImageNet-C Glass blurAUC: 58.78 -> 70.18Improved
Replace TCP confidence loss with steep slope lossViT trustworthiness oracle · ImageNet-C Motion blurAUC: 58.36 -> 69.52Improved
Replace TCP confidence loss with steep slope lossViT trustworthiness oracle · ImageNet-C Zoom blurAUC: 62.47 -> 70.94Improved
Replace TCP† fine-tuning loss with steep slope lossTrustworthiness predictor · MNISTFPR-95%-TPR: 3.33 -> 2.22Improved
Replace TCP† fine-tuning loss with steep slope lossTrustworthiness predictor · CIFAR-10FPR-95%-TPR: 45.07 -> 44.69Improved

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