Open Source Mathematics Olympiad Textbook Collection

Featuring chunked content from 12 open-source mathematics textbooks, including works like 'An Infinitely Large Napkin' and 'Mathematical Reasoning: Writing and Proof'. It is intended for retrieval-augmented generation, embedding, and math reasoning research. The source code for the data pipeline is publicly available on GitHub.

Use Cases

• Fine-tune or evaluate retrieval models on chunked textbook content for mathematical question answering.
• Train embedding models on specialized mathematical text from sources like 'Exploring Combinatorial Mathematics'.
• Benchmark reasoning models using structured textbook content from authors such as Evan Chen and Ted Sundstrom.
• Analyze the linguistic and conceptual structure of advanced mathematics textbooks for educational technology.

Strengths

• Content sourced from 12 distinct open-source mathematics textbooks.
• Includes works from notable authors like Evan Chen and Ted Sundstrom.
• Explicitly intended for retrieval-augmented generation and math reasoning research.

Limitations

• The specific chunking methodology, text quality, and potential duplication are not described.
• No information on the total volume of text (e.g., word count, token count) is provided.
• Limited to the 12 listed textbooks, which may not cover all mathematical olympiad topics.

Provenance

Source

GitHub repositories of the 12 listed open-source textbooks.

Collection Method

Content was chunked using a pipeline with source code at github.com/yoonholee/olympiad-books-open-source-pipeline.

Freshness

Last updated on 2026-02-25.