Language Models Are Capable of Metacognitive Monitoring and Control of Their Internal Activations

Re-assign community