ConfidenceBench

What is ConfidenceBench?

A calibration benchmark evaluating verbalized confidence estimates in frontier LLMs using Brier scores across 200 multiple-choice questions in four categories…

Released

2026-07-10

Evaluates

General AI, Language & Knowledge, cs.AI

Openness

unknown

Importer

Claire Radar

Review status

unreviewed

Reported scores

0

Source provenance

• Original evidence https://arxiv.org/abs/2607.20526

ConfidenceBench paper

• Paper https://arxiv.org/abs/2607.20526

Which sources cite ConfidenceBench?

• ConfidenceBench

Related General AI, Language & Knowledge, cs.AI benchmarks

• $\tau^\tau$-Bench: An Environment for End-To-End, Realistic Agent Construction
• 3DCodeBench
• 5-Dialects-BN
• A Large-Scale Dataset and Benchmark
• A3S-Bench
• AARRI-Bench
• ABC-Bench
• ActBench