bioRxiv · 10.1101/2025.06.02.657538
Genome-Bench: A Scientific Reasoning Benchmark from Real-World Expert Discussions
Abstract
In this short report, we present an automated pipeline tailored for the genomics domain and introduce Genome-Bench, a new benchmark constructed from over a decade of scientific forum discussions on genome engineering. Our pipeline transforms raw interactions into a reinforcement learningfriendly multiple-choice questions format, supported by 3000+ high-quality questionanswer pairs spanning foundational biology, experimental troubleshooting, tool usage, and beyond. To our knowledge, this is the first end-to-end pipeline for teaching LLMs to reason from scientific discussions, with promising potential for generalization across scientific domains beyond biology. The dataset is available at https://huggingface.co/datasets/Mingyin0312/Genome-Bench.
Explore related subjects
Keep this discovery
Explore connections, maps & timelines
Yin, M., Qu, Y., Liu, D., Yang, L., Cong, L., Wang, M.. 2025-06-05. Genome-Bench: A Scientific Reasoning Benchmark from Real-World Expert Discussions. https://doi.org/10.1101/2025.06.02.657538
Cite the original work for its findings. Save a collection to share your selection of sources.