Assistant Professor, New York University
1 paper at NeurIPS 2025
This paper demonstrates that language models possess metacognitive-like abilities to monitor and control their internal neural activations.
