Confidence Calibration in Large Language Models
Research finds that large language models (LLMs) exhibit human-like calibration biases: overconfidence on hard tasks and underconfidence on easy ones. The authors introduce LifeEval, a benchmark for evaluating calibration across difficulty levels.
Article intelligence
Key points
- LLMs are on average overconfident, with confidence exceeding accuracy
- A hard-easy effect is observed: overconfidence on difficult tests, underconfidence on easy tests
- LifeEval is developed to assess calibration across varying difficulty levels
- Findings have implications for the reliability of AI systems
Why it matters
This matters because lLMs are on average overconfident, with confidence exceeding accuracy.
Technical impact
May affect model selection, inference cost, product capability, and evaluation benchmarks.
[2605.23909] Confidence Calibration in Large Language Models
[Submitted on 3 Apr 2026]
Title:Confidence Calibration in Large Language Models
View a PDF of the paper titled Confidence Calibration in Large Language Models, by Noam Michael and 3 other authors
View PDF HTML (experimental)
Abstract:We investigate the calibration of large language models' (LLMs') confidence across diverse tasks. The results of our preregistered study show that the current crop of LLMs are, like people, too sure they are right: confidence exceeds accuracy, on average. Importantly, however, this tendency is moderated by a powerful hard-easy effect, wherein overconfidence is greatest on difficult tests; by contrast, easy tests actually show substantial underconfidence. We develop LifeEval, a test for evaluating model calibration across levels of difficulty.
Subjects:
Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
Cite as: arXiv:2605.23909 [cs.AI]
(or arXiv:2605.23909v1 [cs.AI] for this version)
https://doi.org/10.48550/arXiv.2605.23909
arXiv-issued DOI via DataCite
Submission history
From: Noam Michael [view email] [v1] Fri, 3 Apr 2026 19:43:24 UTC (3,200 KB)
Full-text links:
Access Paper:
View a PDF of the paper titled Confidence Calibration in Large Language Models, by Noam Michael and 3 other authors
View PDF
HTML (experimental)
TeX Source
view license
Current browse context:
cs.AI
new | recent | 2026-05
Change to browse by:
cs cs.LG
References & Citations
NASA ADS
Google Scholar
Semantic Scholar
Loading...
Data provided by:
Bibliographic Tools
Bibliographic and Citation Tools
Bibliographic Explorer Toggle
Bibliographic Explorer (What is the Explorer?)
Connected Papers Toggle
Connected Papers (What is Connected Papers?)
Litmaps Toggle
Litmaps (What is Litmaps?)
scite.ai Toggle
scite Smart Citations (What are Smart Citations?)
Code, Data, Media
Code, Data and Media Associated with this Article
alphaXiv Toggle
alphaXiv (What is alphaXiv?)
Links to Code Toggle
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub Toggle
DagsHub (What is DagsHub?)
GotitPub Toggle
Gotit.pub (What is GotitPub?)
Huggingface Toggle
Hugging Face (What is Huggingface?)
ScienceCast Toggle
ScienceCast (What is ScienceCast?)
Demos
Demos
Replicate Toggle
Replicate (What is Replicate?)
Spaces Toggle
Hugging Face Spaces (What is Spaces?)
Spaces Toggle
TXYZ.AI (What is TXYZ.AI?)
Related Papers
Recommenders and Search Tools
Link to Influence Flower
Influence Flower (What are Influence Flowers?)
Core recommender toggle
CORE Recommender (What is CORE?)
Author
Venue
Institution
Topic
About arXivLabs
arXivLabs: experimental projects with community collaborators
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.
Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)