User: Guest  Login
Title:

A Minimal Model for Compositional Generalization on gSCAN

Document type:
Konferenzbeitrag
Author(s):
Hein, Alice; Diepold, Klaus
Pages contribution:
1-15
Abstract:
Whether neural networks are capable of compositional generalization has been a topic of much debate. Most previous studies on this subject investigate the generalization capabilities of state-of-the-art deep learning architectures. We here take a more bottom-up approach and design a minimal model that displays generalization on a compositional benchmark, namely, the gSCAN dataset. The model is a hybrid architecture that combines layers trained with gradient descent and a selective attention mech...     »
Editor:
Association for Computational Linguistics
Book / Congress title:
Proceedings of the Fifth BlackboxNLP Workshop on Analyzing and Interpreting Neural Networks for NLP
Congress (additional information):
Abu Dhabi, United Arab Emirates (Hybrid)
Year:
2022
Year / month:
2022-12
Month:
Dec
Pages:
15
Reviewed:
ja
Language:
en
Publication format:
WWW
WWW:
https://preview.aclanthology.org/emnlp-22-ingestion/2022.blackboxnlp-1.1.pdf
 BibTeX