On the Self-awareness of Large Reasoning Models' Capability Boundaries

Open in new window