AI.info

Tianli Li

Explore Tianli Li on AI.info.

• When Derived Measurements Mislead: Quantifying and Mitigating LLM Over-Trust with Privileged-Modality Reliability Evidence