Modelpedia
Work in progress
About
Findings
Models
Concepts
Methods
Datasets
Sources
Light
Dark
Temperature Scaling
anchor
Findings
IC-295
The ViT model's ECE can be reduced to near-zero by trivial mean-replacement recalibration while maintaining test accuracy, but NLL increases from 65.35 to 144.66, demonstrating that ECE and accuracy alone are an insufficient reporting standard for calibration
[compared-to]
IC-559
Contextual calibration is most effective for prompt classification while temperature scaling is more effective for response classification, but no single post-hoc method fully resolves miscalibration
[builds-on]
IC-559
Contextual calibration is most effective for prompt classification while temperature scaling is more effective for response classification, but no single post-hoc method fully resolves miscalibration
[primary]