Improving Small Language Models' Mathematical Reasoning via Mix Thoughts Distillation Paper • 2401.11864 • Published Jan 22 • 2