analiza aiqbench
This research exposes a fundamental misalignment between how Large Language Models encode verbal confidence markers versus human metacognitive standards, challenging the reliability of current uncertainty quantification methods. By introducing an optimization-based algorithm that maps specific phrases like "very likely" to actual probability distributions, the study reveals systematic disparities in how AI models interpret and express knowledge boundaries compared to humans. These findings are critical for developers aiming to build trustworthy AI systems where verbal confidence accurately reflects underlying model certainty rather than arbitrary likelihood scores.
"very likely" Means "uncertain"? How LLMs Diverge from Humans in Linguistic Uncertainty Quantification
arXiv:2610.00083v1 Announce Type: new Abstract: Humans express uncertainty verbally via markers (e.g., "possible," "likely"), yet most LLM uncertainty quantification (UQ) relies on costing likelihood- or consistency-based signals. From a cognitive perspective, accurate verbal uncertainty reflects metacognitive monitoring, representing knowledge boundaries ("knowing that you don't know") to support regulation and information seeking. In this paper, we investigate how LLMs diverge from humans in verbal uncertainty quantification and whether verbal markers can reliably quantify LLM uncertainty. We curate a corpus of human uncertainty markers from psychology and decision-science literature and benchmark LLMs against it. We observe that LLMs encode verbal uncertainty with numerical levels that differ substantially from those of humans. We then introduce METHODNAME, a novel optimization-based algorithm that learns an optimal uncertainty profile over uncertainty markers directly from LLM outputs. By fitting a marker-uncertainty mapping to best explain empirical correctness, METHODNAME discovers how much probability mass each verbal marker should convey, rather than estimating uncertainty via repeated sampling. METHODNAME enables a direct, marker-level comparison of confidence semantics between humans and LLMs, disentangling mismatch and revealing systematic confidence disparities in verbal expressions.
Title:"very likely" Means "uncertain"? How LLMs Diverge from Humans in Linguistic Uncertainty Quantification
Abstract:Humans express uncertainty verbally via markers (e.g., "possible," "likely"), yet most LLM uncertainty quantification (UQ) relies on costing likelihood- or consistency-based signals. From a cognitive perspective, accurate verbal uncertainty reflects metacognitive monitoring, representing knowledge boundaries ("knowing that you don't know") to support regulation and information seeking. In this paper, we investigate how LLMs diverge from humans in verbal uncertainty quantification and whether verbal markers can reliably quantify LLM uncertainty. We curate a corpus of human uncertainty markers from psychology and decision-science literature and benchmark LLMs against it. We observe that LLMs encode verbal uncertainty with numerical levels that differ substantially from those of humans. We then introduce METHODNAME, a novel optimization-based algorithm that learns an optimal uncertainty profile over uncertainty markers directly from LLM outputs. By fitting a marker-uncertainty mapping to best explain empirical correctness, METHODNAME discovers how much probability mass each verbal marker should convey, rather than estimating uncertainty via repeated sampling. METHODNAME enables a direct, marker-level comparison of confidence semantics between humans and LLMs, disentangling mismatch and revealing systematic confidence disparities in verbal expressions.
Machine Learning (cs.LG)
arXiv:2610.00083 [cs.LG]
(or arXiv:2610.00083v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2610.00083 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 281 cuvinte
Subiecte conexe
Links #5: 2026/07
LessWrong · 31 iul. 2026
State of the Field: AI for Epistemics and Coordination
LessWrong · 2 oct. 2026
The J-Space Debate, Agent Swarms, and Pacing Frontier AI - Digital Minds Newsletter #4
LessWrong · 18 sept. 2026
Angles of attack for continual learning safety
LessWrong · 16 iun. 2026