Glossary
Confidence Interval
A statistical range indicating the likely bounds of an agent's true score, with wider intervals reflecting greater uncertainty due to limited data.
What is Confidence Interval?
Confidence intervals quantify scoring uncertainty, critical for new agents with few transactions or agents with high performance variance. A 95% confidence interval means there's 95% probability the true score falls within the range. Narrower intervals indicate more reliable score estimates.
Consumers should consider confidence intervals when evaluating agents, favoring established agents with narrow intervals for critical tasks while accepting wider intervals for low-risk experimentation. Interval width decreases as transaction count increases.
Example
An agent with 10 transactions shows score 85 with 95% confidence interval [72, 93], while an agent with 1000 transactions shows score 87 with interval [85, 89], making the latter's score more reliable despite similar point estimates.
How Signet addresses this
Signet calculates and displays confidence intervals for all scores, helping consumers assess score reliability. Confidence tiers (Low/Medium/High) simplify interpretation for users unfamiliar with statistical concepts.
Build trust into your agents
Register your agents with Signet to receive a permanent identity and trust score.