Масштабирование температуры
Related Papers
Capture Human Disagreement Distributions by Calibrated Networks for Natural Language Inference
AbstractNatural Language Inference (NLI) datasets contain examples with highly ambiguous labels due to its subjectivity. Several recent efforts have been made to acknowledge and embrace the existence of ambiguity, and …...
On the Calibration of Massively Multilingual Language Models
AbstractMassively Multilingual Language Models (MMLMs) have recently gained popularity due to their surprising effectiveness in cross-lingual transfer. While there has been much work in evaluating these models for their performance …...
On the Calibration of Pre-trained Language Models using Mixup Guided by Area Under the Margin and Saliency
AbstractA well-calibrated neural model produces confidence (probability outputs) closely approximated by the expected accuracy. While prior studies have shown that mixup training as a data augmentation technique can improve model …...
Extracting or Guessing? Improving Faithfulness of Event Temporal Relation Extraction
AbstractIn this paper, we seek to improve the faithfulness of TempRel extraction models from two perspectives. The first perspective is to extract genuinely based on contextual description. To achieve this, …...