@inproceedings{sanchez-carmona-etal-2024-multilevel-analysis,
title = "Multilevel Analysis of Biomedical Domain Adaptation of Llama 2: What Matters the Most? A Case Study",
author = "Sanchez Carmona, Vicente Ivan and
Jiang, Shanshan and
Suzuki, Takeshi and
Dong, Bin",
editor = "Demner-Fushman, Dina and
Ananiadou, Sophia and
Miwa, Makoto and
Roberts, Kirk and
Tsujii, Junichi",
booktitle = "Proceedings of the 23rd Workshop on Biomedical Natural Language Processing",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.bionlp-1.36",
doi = "10.18653/v1/2024.bionlp-1.36",
pages = "449--456",
abstract = "Domain adaptation of Large Language Models (LLMs) leads to models better suited for a particular domain by capturing patterns from domain text which leads to improvements in downstream tasks. To the naked eye, these improvements are visible; however, the patterns are not so. How can we know which patterns and how much they contribute to changes in downstream scores? Through a Multilevel Analysis we discover and quantify the effect of text patterns on downstream scores of domain-adapted Llama 2 for the task of sentence similarity (BIOSSES dataset). We show that text patterns from PubMed abstracts such as clear writing and simplicity, as well as the amount of biomedical information, are the key for improving downstream scores. Also, we show how another factor not usually quantified contributes equally to downstream scores: choice of hyperparameters for both domain adaptation and fine-tuning.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="sanchez-carmona-etal-2024-multilevel-analysis">
<titleInfo>
<title>Multilevel Analysis of Biomedical Domain Adaptation of Llama 2: What Matters the Most? A Case Study</title>
</titleInfo>
<name type="personal">
<namePart type="given">Vicente</namePart>
<namePart type="given">Ivan</namePart>
<namePart type="family">Sanchez Carmona</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shanshan</namePart>
<namePart type="family">Jiang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Takeshi</namePart>
<namePart type="family">Suzuki</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Bin</namePart>
<namePart type="family">Dong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2024-08</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 23rd Workshop on Biomedical Natural Language Processing</title>
</titleInfo>
<name type="personal">
<namePart type="given">Dina</namePart>
<namePart type="family">Demner-Fushman</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sophia</namePart>
<namePart type="family">Ananiadou</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Makoto</namePart>
<namePart type="family">Miwa</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kirk</namePart>
<namePart type="family">Roberts</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Junichi</namePart>
<namePart type="family">Tsujii</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Bangkok, Thailand</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Domain adaptation of Large Language Models (LLMs) leads to models better suited for a particular domain by capturing patterns from domain text which leads to improvements in downstream tasks. To the naked eye, these improvements are visible; however, the patterns are not so. How can we know which patterns and how much they contribute to changes in downstream scores? Through a Multilevel Analysis we discover and quantify the effect of text patterns on downstream scores of domain-adapted Llama 2 for the task of sentence similarity (BIOSSES dataset). We show that text patterns from PubMed abstracts such as clear writing and simplicity, as well as the amount of biomedical information, are the key for improving downstream scores. Also, we show how another factor not usually quantified contributes equally to downstream scores: choice of hyperparameters for both domain adaptation and fine-tuning.</abstract>
<identifier type="citekey">sanchez-carmona-etal-2024-multilevel-analysis</identifier>
<identifier type="doi">10.18653/v1/2024.bionlp-1.36</identifier>
<location>
<url>https://aclanthology.org/2024.bionlp-1.36</url>
</location>
<part>
<date>2024-08</date>
<extent unit="page">
<start>449</start>
<end>456</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Multilevel Analysis of Biomedical Domain Adaptation of Llama 2: What Matters the Most? A Case Study
%A Sanchez Carmona, Vicente Ivan
%A Jiang, Shanshan
%A Suzuki, Takeshi
%A Dong, Bin
%Y Demner-Fushman, Dina
%Y Ananiadou, Sophia
%Y Miwa, Makoto
%Y Roberts, Kirk
%Y Tsujii, Junichi
%S Proceedings of the 23rd Workshop on Biomedical Natural Language Processing
%D 2024
%8 August
%I Association for Computational Linguistics
%C Bangkok, Thailand
%F sanchez-carmona-etal-2024-multilevel-analysis
%X Domain adaptation of Large Language Models (LLMs) leads to models better suited for a particular domain by capturing patterns from domain text which leads to improvements in downstream tasks. To the naked eye, these improvements are visible; however, the patterns are not so. How can we know which patterns and how much they contribute to changes in downstream scores? Through a Multilevel Analysis we discover and quantify the effect of text patterns on downstream scores of domain-adapted Llama 2 for the task of sentence similarity (BIOSSES dataset). We show that text patterns from PubMed abstracts such as clear writing and simplicity, as well as the amount of biomedical information, are the key for improving downstream scores. Also, we show how another factor not usually quantified contributes equally to downstream scores: choice of hyperparameters for both domain adaptation and fine-tuning.
%R 10.18653/v1/2024.bionlp-1.36
%U https://aclanthology.org/2024.bionlp-1.36
%U https://doi.org/10.18653/v1/2024.bionlp-1.36
%P 449-456
Markdown (Informal)
[Multilevel Analysis of Biomedical Domain Adaptation of Llama 2: What Matters the Most? A Case Study](https://aclanthology.org/2024.bionlp-1.36) (Sanchez Carmona et al., BioNLP-WS 2024)
ACL