CORTEXA
← Browse
arxivcs.AIcs.LGstat.ML2026-07-10

ConfidenceBench: Evaluating Confidence Calibration in Large Language Models

Matthew ffrench-Constant, Daniel Yang, Xinmeng Huang, Sanyam Kapoor

Large language models (LLMs) are increasingly deployed in settings where fluent but incorrect answers can be costly. In these settings, accuracy alone is insufficient: models must also know when they are likely to be wrong. We present ConfidenceBench, a calibration benchmark that evaluates verbalized confidence estimates in 15 frontier LLMs using the Brier score, a proper scoring rule that incentivises truthful probability reporting. Confidence is elicited via prompting, requiring no access to model logits and making the framework applicable to both closed-source and open-source systems. The benchmark comprises 200 private multiple-choice questions across four categories: spatial reasoning, high-precision mathematics, word lookup, and unknowable questions. Across three independent runs, Claude Opus 4.6 and Gemini 3.1 Pro Preview achieve the lowest Brier scores, both reported as 0.103. Both substantially outperform the calibrated-random baseline of 0.1875, while Gemini 3.1 Flash-Lite scores 0.367, indicating severe miscalibration. Accuracy and calibration diverge substantially across model families: the most accurate model is not the best-calibrated, and several models perform worse than the calibrated-random Brier baseline despite reasonable accuracy. These results show that verbalized confidence calibration is a distinct and practically important axis of LLM reliability, complementary to standard accuracy-based evaluation.

View free PDFSource page

Related papers

arxivcs.LGcs.AIcs.CLstat.ML2026-06-26

What LLMs explain is not what they believe: Evaluating explanation sufficiency under models' own input beliefs

Nhi Nguyen, Shauli Ravfogel, Rajesh Ranganath

Large language models (LLMs) are increasingly deployed in high-stakes domains, where free-text explanations such as chain-of-thought and post-hoc rationales are used to justify model outputs. Yet it remains unclear whether these explanations are sufficient, i.e., if they contain…

View free PDFSource page
arxivcs.LGcs.AIstat.ML2026-06-27

Uncertainty-Aware Sequential Decision Rules for Event-Triggered LLM Invocation in Streaming Systems

Zhaohui Wang

Streaming inference pipelines increasingly pair lightweight fast models with Large Language Models (LLMs) that provide rich semantic understanding at substantial cost. The central question of when to invoke the LLM has received limited formal treatment. We cast this as a risk-bas…

View free PDFSource page
arxivcs.CLcs.AIcs.LGstat.ML2026-07-05

Telescope: Improving Zero Shot Detection of LLM Generated Content By Measuring Token Repetition Probability

Christopher Nassif, Josh F. Cooper

Distinguishing Large Language Model (LLM) generated text from human writing is a critical and difficult challenge. While LLMs are trained to write like humans, we hypothesize that this training leaves an indelible mark. LLMs develop a particularly strong aversion to token repetit…

View free PDFSource page
arxivcs.LGcs.AIstat.ML2026-07-09

Reinforcing the Generation Order of Multimodal Masked Diffusion Models

Yidong Ouyang, Zhe Wang, Sourav Bhabesh, Dmitriy Bespalov

Diffusion Language Models (DLMs) have recently achieved substantial progress in natural language generation tasks. Recent research demonstrates that adaptive token generation ordering can significantly improve performance in mathematical reasoning and code synthesis applications.…

View free PDFSource page
arxivstat.MLcs.AIcs.LGecon.EM2026-07-24

CausalForge: A Formally Grounded, Self-Improving Agentic Framework for Automated Research in Causal Inference

Jiyuan Tan, Vasilis Syrgkanis

Automating theoretical research is constrained not only by the generation of candidate results, but also by their reliable evaluation. A common approach is to close the research loop with a large language model (LLM) reviewer. However, such reviewers remain empirically unreliable…

View free PDFSource page
arxivcs.LGcs.AIstat.ML2026-06-29

Behavior Cloning is Not All You Need: The Optimality of On-Policy Distillation for Noisy Expert Feedback

Ved Sriraman, Peihan Liu, Daniel Hsu, Adam Block

Imitation Learning is a natural framework for learning in sequential decision-making systems and has emerged as the dominant paradigm through which we understand language model training. A central puzzle is that, while in theory offline IL can be horizon-free and optimal, in prac…

View free PDFSource page