Lon-ea at SemEval-2023 Task 11: A Comparison of Activation Functions for Soft and Hard Label Prediction

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Hosseini, Peyman, Hosseini, Mehran, Al-Azzawi, Sana Sabah, Liwicki, Marcus, Castro, Ignacio, Purver, Matthew
Format: Preprint
Veröffentlicht: 2023
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866910286830632960
author Hosseini, Peyman
Hosseini, Mehran
Al-Azzawi, Sana Sabah
Liwicki, Marcus
Castro, Ignacio
Purver, Matthew
author_facet Hosseini, Peyman
Hosseini, Mehran
Al-Azzawi, Sana Sabah
Liwicki, Marcus
Castro, Ignacio
Purver, Matthew
contents We study the influence of different activation functions in the output layer of deep neural network models for soft and hard label prediction in the learning with disagreement task. In this task, the goal is to quantify the amount of disagreement via predicting soft labels. To predict the soft labels, we use BERT-based preprocessors and encoders and vary the activation function used in the output layer, while keeping other parameters constant. The soft labels are then used for the hard label prediction. The activation functions considered are sigmoid as well as a step-function that is added to the model post-training and a sinusoidal activation function, which is introduced for the first time in this paper.
format Preprint
id arxiv_https___arxiv_org_abs_2303_02468
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Lon-ea at SemEval-2023 Task 11: A Comparison of Activation Functions for Soft and Hard Label Prediction
Hosseini, Peyman
Hosseini, Mehran
Al-Azzawi, Sana Sabah
Liwicki, Marcus
Castro, Ignacio
Purver, Matthew
Computation and Language
Machine Learning
I.2.7
We study the influence of different activation functions in the output layer of deep neural network models for soft and hard label prediction in the learning with disagreement task. In this task, the goal is to quantify the amount of disagreement via predicting soft labels. To predict the soft labels, we use BERT-based preprocessors and encoders and vary the activation function used in the output layer, while keeping other parameters constant. The soft labels are then used for the hard label prediction. The activation functions considered are sigmoid as well as a step-function that is added to the model post-training and a sinusoidal activation function, which is introduced for the first time in this paper.
title Lon-ea at SemEval-2023 Task 11: A Comparison of Activation Functions for Soft and Hard Label Prediction
topic Computation and Language
Machine Learning
I.2.7
url https://arxiv.org/abs/2303.02468