BioAlchemy distills biology literature into 345K verifiable reasoning problems, enabling an 8B model to improve 9.12% on biology benchmarks through topic-aligned reinforcement learning.
The correct answer must be choice number {target_correct_position}
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
BioAlchemy: Distilling Biological Literature into Reasoning-Ready Reinforcement Learning Training Data
BioAlchemy distills biology literature into 345K verifiable reasoning problems, enabling an 8B model to improve 9.12% on biology benchmarks through topic-aligned reinforcement learning.