TY  - RPRT
TI  - Closing the Confidence-Faithfulness Gap in Large Language Models
AU  - Miranda Muqing Miao
AU  - Lyle Ungar
PY  - 2026
UR  - https://arxiv.org/abs/2603.25052
ID  - 2603.25052
ER  -