# Class-Similarity Based Label Smoothing for Confidence Calibration

arXiv: https://arxiv.org/abs/2006.14028 · HTML: https://mlsanity.com/paper/2006.14028/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 24 Jun 2020.

## Built on by (most cited first, 3 of 3)

- [ConGraT: Self-Supervised Contrastive Pretraining for Joint Graph and Text Embeddings](https://mlsanity.com/paper/2305.14321/) (2023)
- [Teaching Metric Distance to Discrete Autoregressive Language Models](https://mlsanity.com/paper/2503.02379/) (2025)
- [From Discord to Harmony: Decomposed Consonance-based Training for Improved Audio Chord Estimation](https://mlsanity.com/paper/2509.01588/) (2025)

## Abstract (arXiv)

Generating confidence calibrated outputs is of utmost importance for the applications of deep neural networks in safety-critical decision-making systems. The output of a neural network is a probability distribution where the scores are estimated confidences of the input belonging to the corresponding classes, and hence they represent a complete estimate of the output likelihood relative to all classes. In this paper, we propose a novel form of label smoothing to improve confidence calibration. Since different classes are of different intrinsic similarities, more similar classes should result in closer probability values in the final output. This motivates the development of a new smooth label where the label values are based on similarities with the reference class. We adopt different similarity measurements, including those that capture feature-based similarities or semantic similarity. We demonstrate through extensive experiments, on various datasets and network architectures, that our approach consistently outperforms state-of-the-art calibration techniques including uniform label smoothing.