Announcement_6

Our paper “A Positive Case for Faithfulness: LLM Self-Explanations Help Predict Model Behavior” was accepted to ICML 2026 :kr:.