Active Curriculum Language Modeling over a Hybrid Pre-training Method

Eleni Fysikoudi; Sharid Loáiciga; Asad Sayeed

Active Curriculum Language Modeling over a Hybrid Pre-training Method

Eleni Fysikoudi, Sharid Loáiciga, Asad B. Sayeed

Correct Metadata for

Important: The Anthology treat PDFs as authoritative. Please use this form only to correct data that is out of line with the PDF. See our corrections guidelines if you need to change the PDF.

Title Adjust the title. Retain tags such as <fixed-case>.

Authors Adjust author names and order to match the PDF.

Abstract Correct abstract if needed. Retain XML formatting tags such as <tex-math>.

Verification against PDF Ensure that the new title/authors match the snapshot below. (If there is no snapshot or it is too small, consult the PDF.)

Authors concatenated from the text boxes above:

ALL author names match the snapshot above—including middle initials, hyphens, and accents.

Abstract

We apply the Active Curriculum Language Modeling (ACLM) method to the constrained pretraining setting of the 2025 BabyLM Challenge, where models are limited by both data and compute budgets. Using GPT-BERT (Charpentier and Samuel, 2024) as the base architecture, we investigate the impact of surprisal-based example selection for constructing a training curriculum. In addition, we conduct a targeted hyperparameter search over tokenizer size and batch size. Our approach yields stable pretrained models that surpass the official baseline on multiple evaluation tasks, demonstrating ACLM’s potential for improving performance and generalization in low-resource pretraining scenarios.

Anthology ID:: 2025.babylm-main.34
Volume:: Proceedings of the First BabyLM Workshop
Month:: November
Year:: 2025
Address:: Suzhou, China
Editors:: Lucas Charpentier, Leshem Choshen, Ryan Cotterell, Mustafa Omer Gul, Michael Y. Hu, Jing Liu, Jaap Jumelet, Tal Linzen, Aaron Mueller, Candace Ross, Raj Sanjay Shah, Alex Warstadt, Ethan Gotlieb Wilcox, Adina Williams
Venue:: BabyLM
SIG:
Publisher:: Association for Computational Linguistics
Note:
Pages:: 488–495
Language:
URL:: https://aclanthology.org/2025.babylm-main.34/
DOI:
Bibkey:
Cite (ACL):: Eleni Fysikoudi, Sharid Loáiciga, and Asad B. Sayeed. 2025. Active Curriculum Language Modeling over a Hybrid Pre-training Method. In Proceedings of the First BabyLM Workshop, pages 488–495, Suzhou, China. Association for Computational Linguistics.
Cite (Informal):: Active Curriculum Language Modeling over a Hybrid Pre-training Method (Fysikoudi et al., BabyLM 2025)
Copy Citation:
PDF:: https://aclanthology.org/2025.babylm-main.34.pdf

PDF Cite Search Fix data

Export citation

BibTeX
MODS XML
Endnote
Preformatted

@inproceedings{fysikoudi-etal-2025-active,
    title = "Active Curriculum Language Modeling over a Hybrid Pre-training Method",
    author = "Fysikoudi, Eleni  and
      Lo{\'a}iciga, Sharid  and
      Sayeed, Asad B.",
    editor = "Charpentier, Lucas  and
      Choshen, Leshem  and
      Cotterell, Ryan  and
      Gul, Mustafa Omer  and
      Hu, Michael Y.  and
      Liu, Jing  and
      Jumelet, Jaap  and
      Linzen, Tal  and
      Mueller, Aaron  and
      Ross, Candace  and
      Shah, Raj Sanjay  and
      Warstadt, Alex  and
      Wilcox, Ethan Gotlieb  and
      Williams, Adina",
    booktitle = "Proceedings of the First BabyLM Workshop",
    month = nov,
    year = "2025",
    address = "Suzhou, China",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2025.babylm-main.34/",
    pages = "488--495",
    ISBN = "TODO",
    abstract = "We apply the Active Curriculum Language Modeling (ACLM) method to the constrained pretraining setting of the 2025 BabyLM Challenge, where models are limited by both data and compute budgets. Using GPT-BERT (Charpentier and Samuel, 2024) as the base architecture, we investigate the impact of surprisal-based example selection for constructing a training curriculum. In addition, we conduct a targeted hyperparameter search over tokenizer size and batch size. Our approach yields stable pretrained models that surpass the official baseline on multiple evaluation tasks, demonstrating ACLM{'}s potential for improving performance and generalization in low-resource pretraining scenarios."
}

Download as File

<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="fysikoudi-etal-2025-active">
    <titleInfo>
        <title>Active Curriculum Language Modeling over a Hybrid Pre-training Method</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Eleni</namePart>
        <namePart type="family">Fysikoudi</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Sharid</namePart>
        <namePart type="family">Loáiciga</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Asad</namePart>
        <namePart type="given">B</namePart>
        <namePart type="family">Sayeed</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2025-11</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Proceedings of the First BabyLM Workshop</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Lucas</namePart>
            <namePart type="family">Charpentier</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Leshem</namePart>
            <namePart type="family">Choshen</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Ryan</namePart>
            <namePart type="family">Cotterell</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Mustafa</namePart>
            <namePart type="given">Omer</namePart>
            <namePart type="family">Gul</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Michael</namePart>
            <namePart type="given">Y</namePart>
            <namePart type="family">Hu</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Jing</namePart>
            <namePart type="family">Liu</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Jaap</namePart>
            <namePart type="family">Jumelet</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Tal</namePart>
            <namePart type="family">Linzen</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Aaron</namePart>
            <namePart type="family">Mueller</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Candace</namePart>
            <namePart type="family">Ross</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Raj</namePart>
            <namePart type="given">Sanjay</namePart>
            <namePart type="family">Shah</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Alex</namePart>
            <namePart type="family">Warstadt</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Ethan</namePart>
            <namePart type="given">Gotlieb</namePart>
            <namePart type="family">Wilcox</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Adina</namePart>
            <namePart type="family">Williams</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Suzhou, China</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
        <identifier type="isbn">TODO</identifier>
    </relatedItem>
    <abstract>We apply the Active Curriculum Language Modeling (ACLM) method to the constrained pretraining setting of the 2025 BabyLM Challenge, where models are limited by both data and compute budgets. Using GPT-BERT (Charpentier and Samuel, 2024) as the base architecture, we investigate the impact of surprisal-based example selection for constructing a training curriculum. In addition, we conduct a targeted hyperparameter search over tokenizer size and batch size. Our approach yields stable pretrained models that surpass the official baseline on multiple evaluation tasks, demonstrating ACLM’s potential for improving performance and generalization in low-resource pretraining scenarios.</abstract>
    <identifier type="citekey">fysikoudi-etal-2025-active</identifier>
    <location>
        <url>https://aclanthology.org/2025.babylm-main.34/</url>
    </location>
    <part>
        <date>2025-11</date>
        <extent unit="page">
            <start>488</start>
            <end>495</end>
        </extent>
    </part>
</mods>
</modsCollection>

Download as File

%0 Conference Proceedings
%T Active Curriculum Language Modeling over a Hybrid Pre-training Method
%A Fysikoudi, Eleni
%A Loáiciga, Sharid
%A Sayeed, Asad B.
%Y Charpentier, Lucas
%Y Choshen, Leshem
%Y Cotterell, Ryan
%Y Gul, Mustafa Omer
%Y Hu, Michael Y.
%Y Liu, Jing
%Y Jumelet, Jaap
%Y Linzen, Tal
%Y Mueller, Aaron
%Y Ross, Candace
%Y Shah, Raj Sanjay
%Y Warstadt, Alex
%Y Wilcox, Ethan Gotlieb
%Y Williams, Adina
%S Proceedings of the First BabyLM Workshop
%D 2025
%8 November
%I Association for Computational Linguistics
%C Suzhou, China
%@ TODO
%F fysikoudi-etal-2025-active
%X We apply the Active Curriculum Language Modeling (ACLM) method to the constrained pretraining setting of the 2025 BabyLM Challenge, where models are limited by both data and compute budgets. Using GPT-BERT (Charpentier and Samuel, 2024) as the base architecture, we investigate the impact of surprisal-based example selection for constructing a training curriculum. In addition, we conduct a targeted hyperparameter search over tokenizer size and batch size. Our approach yields stable pretrained models that surpass the official baseline on multiple evaluation tasks, demonstrating ACLM’s potential for improving performance and generalization in low-resource pretraining scenarios.
%U https://aclanthology.org/2025.babylm-main.34/
%P 488-495

Download as File

Markdown (Informal)

[Active Curriculum Language Modeling over a Hybrid Pre-training Method](https://aclanthology.org/2025.babylm-main.34/) (Fysikoudi et al., BabyLM 2025)

Active Curriculum Language Modeling over a Hybrid Pre-training Method (Fysikoudi et al., BabyLM 2025)

ACL

Eleni Fysikoudi, Sharid Loáiciga, and Asad B. Sayeed. 2025. Active Curriculum Language Modeling over a Hybrid Pre-training Method. In Proceedings of the First BabyLM Workshop, pages 488–495, Suzhou, China. Association for Computational Linguistics.