Domain-Adaptive Continued Pre-Training of Small Language Models Paper • 2504.09687 • Published Apr 13, 2025 • 2