BabyBabelLM: A Multilingual Benchmark of Developmentally Plausible Training Data

ACL ARR 2025 July Submission762 Authors

Abstract: We present BabyBabelLM, a multilingual collection of datasets modeling the language a person observes from birth until they acquire a native language. We curate developmentally plausible pretraining data in each of 45 languages, aiming to cover the equivalent of 100M English words of content. We compile evaluation suites and train baseline models in each language. BabyBabelLM aims to facilitate multilingual pretraining and cognitive modeling.

Paper Type: Long

Research Area: Multilingualism and Cross-Lingual NLP

Research Area Keywords: multilingualism, multilingual evaluation, less-resourced languages, resources for less-resourced languages, corpus creation, multilingual corpora

Contribution Types: Model analysis & interpretability, Approaches to low-resource settings, Data resources

Languages Studied: Achinese, Afrikaans, Arabic, Balinese, Buginese, Bulgarian, Czech, Welsh, Danish, German, Modern Greek (1453-), English, Estonian, Basque, Persian, French, Hebrew, Croatian, Hungarian, Indonesian, Icelandic, Italian, Javanese, Japanese, Korean, Makasar, Minangkabau, Dutch, Norwegian, Pedi, Polish, Portuguese, Romanian, Russian, Southern Sotho, Spanish, Serbian, Sundanese, Swedish, Turkish, Ukrainian, Xhosa, Yue Chinese, Chinese, Zulu

Reassignment Request Area Chair: This is not a resubmission

Reassignment Request Reviewers: This is not a resubmission

A1 Limitations Section: This paper has a limitations section.

A2 Potential Risks: No

A2 Elaboration: Our resource is intended to be used in the intersection of language acquisition and low-resource language modelling, and presents no potential risks.

B Use Or Create Scientific Artifacts: Yes

B1 Cite Creators Of Artifacts: Yes

B1 Elaboration: 3.2.1 and Appendix B

B2 Discuss The License For Artifacts: Yes

B2 Elaboration: 3.1.3 and Appendix B

B3 Artifact Use Consistent With Intended Use: Yes

B3 Elaboration: 3.1.3

B4 Data Contains Personally Identifying Info Or Offensive Content: No

B4 Elaboration: 3.3

B5 Documentation Of Artifacts: Yes

B5 Elaboration: 3.2.2 and Appendix B

B6 Statistics For Data: Yes

B6 Elaboration: 3.2.2, Appendix B, Table 2

C Computational Experiments: Yes

C1 Model Size And Budget: Yes

C1 Elaboration: 5

C2 Experimental Setup And Hyperparameters: Yes

C2 Elaboration: 5

C3 Descriptive Statistics: Yes

C3 Elaboration: 5

C4 Parameters For Packages: N/A

D Human Subjects Including Annotators: No

D1 Instructions Given To Participants: N/A

D2 Recruitment And Payment: N/A

D3 Data Consent: Yes

D3 Elaboration: 3.1.3

D4 Ethics Review Board Approval: N/A

D5 Characteristics Of Annotators: N/A

E Ai Assistants In Research Or Writing: No

E1 Information About Use Of Ai Assistants: N/A

Author Submission Checklist: yes

Submission Number: 762