paper-with-me

Papers

Using Focal Loss to Fight Shallow Heuristics: An Empirical Analysis of Modulated Cross-Entropy in Natural Language Inference

2022-11-23 · Frano Rajič, Ivan Stresec, Axel Marmet, Tim Poštuvan

There is no such thing as a perfect dataset. In some datasets, deep neural networks discover underlying heuristics that allow them to take shortcuts in the learning process, resulting in poor generalization capability. Instead of using standard cross-entropy, we explore whether a modulated version of cross-entropy called focal loss can constrain the model so as not to use heuristics and improve generalization performance. Our experiments in natural language inference show that focal loss has a regularizing impact on the learning process, increasing accuracy on out-of-distribution data, but slightly decreasing performance on in-distribution data. Despite the improved out-of-distribution performance, we demonstrate the shortcomings of focal loss and its inferiority in comparison to the performance of methods such as unbiased focal loss and self-debiasing ensembles.

📄 PDF Abstract BibTeX arXiv:2211.13331

Code (1)

m43/focal-loss-against-heuristics 공식 구현 pytorch

Tasks

Natural Language Inference

Methods 이 논문이 사용한 방법론

Focal Loss A Focal Loss function addresses class imbalance during training in tasks like object detection. Focal loss applies a modulating term to the cross entropy loss in order to…

Similar Papers 제목 키워드 기반

AdaFocal: Calibration-aware Adaptive Focal Loss

2022-11-21 · Arindam Ghosh, Thomas Schaaf, Matthew R. Gormley

Much recent work has been devoted to the problem of ensuring that a neural network's confidence scores match the true probability of being correct, i.e. the calibration problem. Of note, it was found that training with f…

Out-of-Distribution Detection

Functional Properties of the Focal-Entropy

2026-03-03 · Jaimin Shah, Martina Cardone, Alex Dytso arxiv

The focal-loss has become a widely used alternative to cross-entropy in class-imbalanced classification problems, particularly in computer vision. Despite its empirical success, a systematic information-theoretic study o…

Revisiting Reweighted Risk for Calibration: AURC, Focal Loss, and Inverse Focal Loss

2025-05-29 · Han Zhou, Sebastian G. Gruber, Teodora Popordanoska, Matthew B. Blaschko

Several variants of reweighted risk functionals, such as focal losss, inverse focal loss, and the Area Under the Risk-Coverage Curve (AURC), have been proposed in the literature and claims have been made in relation to t…

FocalPO: Enhancing Preference Optimizing by Focusing on Correct Preference Rankings

2025-01-11 · Tong Liu, Xiao Yu, Wenxuan Zhou, Jindong Gu 외

Efficient preference optimization algorithms such as Direct Preference Optimization (DPO) have become a popular approach in aligning large language models (LLMs) with human preferences. These algorithms implicitly treat …

Calibrating Deep Neural Networks using Focal Loss

2020-02-21 · NeurIPS 2020 12 · Jishnu Mukhoti, Viveka Kulharia, Amartya Sanyal, Stuart Golodetz 외

Miscalibration - a mismatch between a model's confidence and its correctness - of Deep Neural Networks (DNNs) makes their predictions hard to rely on. Ideally, we want networks to be accurate, calibrated and confident. W…