@inproceedings{abdelali-etal-2022-post,
title = "Post-hoc analysis of {A}rabic transformer models",
author = "Abdelali, Ahmed and
Durrani, Nadir and
Dalvi, Fahim and
Sajjad, Hassan",
editor = "Bastings, Jasmijn and
Belinkov, Yonatan and
Elazar, Yanai and
Hupkes, Dieuwke and
Saphra, Naomi and
Wiegreffe, Sarah",
booktitle = "Proceedings of the Fifth BlackboxNLP Workshop on Analyzing and Interpreting Neural Networks for NLP",
month = dec,
year = "2022",
address = "Abu Dhabi, United Arab Emirates (Hybrid)",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2022.blackboxnlp-1.8",
doi = "10.18653/v1/2022.blackboxnlp-1.8",
pages = "91--103",
abstract = "Arabic is a Semitic language which is widely spoken with many dialects. Given the success of pre-trained language models, many transformer models trained on Arabic and its dialects have surfaced. While there have been an extrinsic evaluation of these models with respect to downstream NLP tasks, no work has been carried out to analyze and compare their internal representations. We probe how linguistic information is encoded in the transformer models, trained on different Arabic dialects. We perform a layer and neuron analysis on the models using morphological tagging tasks for different dialects of Arabic and a dialectal identification task. Our analysis enlightens interesting findings such as: i) word morphology is learned at the lower and middle layers, ii) while syntactic dependencies are predominantly captured at the higher layers, iii) despite a large overlap in their vocabulary, the MSA-based models fail to capture the nuances of Arabic dialects, iv) we found that neurons in embedding layers are polysemous in nature, while the neurons in middle layers are exclusive to specific properties.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="abdelali-etal-2022-post">
<titleInfo>
<title>Post-hoc analysis of Arabic transformer models</title>
</titleInfo>
<name type="personal">
<namePart type="given">Ahmed</namePart>
<namePart type="family">Abdelali</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Nadir</namePart>
<namePart type="family">Durrani</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Fahim</namePart>
<namePart type="family">Dalvi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hassan</namePart>
<namePart type="family">Sajjad</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2022-12</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the Fifth BlackboxNLP Workshop on Analyzing and Interpreting Neural Networks for NLP</title>
</titleInfo>
<name type="personal">
<namePart type="given">Jasmijn</namePart>
<namePart type="family">Bastings</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yonatan</namePart>
<namePart type="family">Belinkov</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yanai</namePart>
<namePart type="family">Elazar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Dieuwke</namePart>
<namePart type="family">Hupkes</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Naomi</namePart>
<namePart type="family">Saphra</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sarah</namePart>
<namePart type="family">Wiegreffe</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Abu Dhabi, United Arab Emirates (Hybrid)</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Arabic is a Semitic language which is widely spoken with many dialects. Given the success of pre-trained language models, many transformer models trained on Arabic and its dialects have surfaced. While there have been an extrinsic evaluation of these models with respect to downstream NLP tasks, no work has been carried out to analyze and compare their internal representations. We probe how linguistic information is encoded in the transformer models, trained on different Arabic dialects. We perform a layer and neuron analysis on the models using morphological tagging tasks for different dialects of Arabic and a dialectal identification task. Our analysis enlightens interesting findings such as: i) word morphology is learned at the lower and middle layers, ii) while syntactic dependencies are predominantly captured at the higher layers, iii) despite a large overlap in their vocabulary, the MSA-based models fail to capture the nuances of Arabic dialects, iv) we found that neurons in embedding layers are polysemous in nature, while the neurons in middle layers are exclusive to specific properties.</abstract>
<identifier type="citekey">abdelali-etal-2022-post</identifier>
<identifier type="doi">10.18653/v1/2022.blackboxnlp-1.8</identifier>
<location>
<url>https://aclanthology.org/2022.blackboxnlp-1.8</url>
</location>
<part>
<date>2022-12</date>
<extent unit="page">
<start>91</start>
<end>103</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Post-hoc analysis of Arabic transformer models
%A Abdelali, Ahmed
%A Durrani, Nadir
%A Dalvi, Fahim
%A Sajjad, Hassan
%Y Bastings, Jasmijn
%Y Belinkov, Yonatan
%Y Elazar, Yanai
%Y Hupkes, Dieuwke
%Y Saphra, Naomi
%Y Wiegreffe, Sarah
%S Proceedings of the Fifth BlackboxNLP Workshop on Analyzing and Interpreting Neural Networks for NLP
%D 2022
%8 December
%I Association for Computational Linguistics
%C Abu Dhabi, United Arab Emirates (Hybrid)
%F abdelali-etal-2022-post
%X Arabic is a Semitic language which is widely spoken with many dialects. Given the success of pre-trained language models, many transformer models trained on Arabic and its dialects have surfaced. While there have been an extrinsic evaluation of these models with respect to downstream NLP tasks, no work has been carried out to analyze and compare their internal representations. We probe how linguistic information is encoded in the transformer models, trained on different Arabic dialects. We perform a layer and neuron analysis on the models using morphological tagging tasks for different dialects of Arabic and a dialectal identification task. Our analysis enlightens interesting findings such as: i) word morphology is learned at the lower and middle layers, ii) while syntactic dependencies are predominantly captured at the higher layers, iii) despite a large overlap in their vocabulary, the MSA-based models fail to capture the nuances of Arabic dialects, iv) we found that neurons in embedding layers are polysemous in nature, while the neurons in middle layers are exclusive to specific properties.
%R 10.18653/v1/2022.blackboxnlp-1.8
%U https://aclanthology.org/2022.blackboxnlp-1.8
%U https://doi.org/10.18653/v1/2022.blackboxnlp-1.8
%P 91-103
Markdown (Informal)
[Post-hoc analysis of Arabic transformer models](https://aclanthology.org/2022.blackboxnlp-1.8) (Abdelali et al., BlackboxNLP 2022)
ACL
- Ahmed Abdelali, Nadir Durrani, Fahim Dalvi, and Hassan Sajjad. 2022. Post-hoc analysis of Arabic transformer models. In Proceedings of the Fifth BlackboxNLP Workshop on Analyzing and Interpreting Neural Networks for NLP, pages 91–103, Abu Dhabi, United Arab Emirates (Hybrid). Association for Computational Linguistics.