Beyond Reasoning Gains: Mitigating General Capabilities Forgetting in Large Reasoning Models Paper • 2510.21978 • Published Oct 24 • 15