conftrace_

overconfidence calibration

2 papers

Explore in graph

Co-occurring keywords · papers mentioning both

ethical alignment (1) document visual question answering (1) atypical sample (1) cross-entropy loss (1) failure detection (1) logit magnitude (1) out-of-distribution detection (1) risk coverage (1) deep neural network (1) confidence calibration (1)

Papers

The Confidence Paradox: Can LLM Know When It’s Wrong? IJCNLP 2025 Typicalness-Aware Learning for Failure Detection NeurIPS 2024