Paper Detail

TheBioCollection: Unified Pre-Training Scale LLM Corpus for Biology

Hyunjin Seo, Hyeon Hwang, Gyubok Lee, Jay Shin, Jimin Park, Taesoo Kim, Sanghoon Lee, Hongjoon Ahn, Sungjun Han, Sangwon Jung

arxiv Score 8.8

Published 2026-07-09 · First seen 2026-07-13

General AI

Abstract

The push toward large language models for biology (BioLM) has created a need for training corpora that can endow models with a genuine understanding of biology. However, existing biological resources, such as molecular databases, protein repositories, genomic annotations, single-cell atlases, and pathway databases, are scattered across heterogeneous formats and remain unorganized into a cohesive corpus for language model training. We present TheBioCollection, a 52.6B-token pre-training-scale corpus that converts these disparate resources into a unified, training-ready form spanning small molecules, proteins, genomic sequences, cells, and pathways. Beyond consolidating existing data, TheBioCollection enriches each record with tool-computed biological properties and introduces new instruction tasks for capabilities that current corpora barely cover. We pair the corpus with TheBioCollection-Eval, a matched suite probing recognition, generation, and prediction across molecular, protein, genomic, cellular, and cross-domain settings. Holding the base Gravity-16B-A3B architecture fixed, training on TheBioCollection more than doubles its overall score on TheBioCollection-Eval with gains in every domain, while leaving general linguistic ability nearly intact.

Workflow Status

Review status
pending
Role
unreviewed
Read priority
soon
Vote
Not set.
Saved
no
Collections
Not filed yet.
Next action
Not filled yet.

Reading Brief

No structured notes yet. Add `summary_sections`, `why_relevant`, `claim_impact`, or `next_action` in `papers.jsonl` to enrich this view.

Why It Surfaced

No ranking explanation is available yet.

Tags

No tags.

BibTeX

@article{seo2026thebiocollection,
  title = {TheBioCollection: Unified Pre-Training Scale LLM Corpus for Biology},
  author = {Hyunjin Seo and Hyeon Hwang and Gyubok Lee and Jay Shin and Jimin Park and Taesoo Kim and Sanghoon Lee and Hongjoon Ahn and Sungjun Han and Sangwon Jung},
  year = {2026},
  abstract = {The push toward large language models for biology (BioLM) has created a need for training corpora that can endow models with a genuine understanding of biology. However, existing biological resources, such as molecular databases, protein repositories, genomic annotations, single-cell atlases, and pathway databases, are scattered across heterogeneous formats and remain unorganized into a cohesive corpus for language model training. We present TheBioCollection, a 52.6B-token pre-training-scale cor},
  url = {https://arxiv.org/abs/2607.08803},
  keywords = {q-bio.QM, cs.AI, cs.LG},
  eprint = {2607.08803},
  archiveprefix = {arXiv},
}

Metadata

{}