Unverified50% confidenceBenchmarkExact time
When given input-irrelevant junk messages, the model's maximum confidence drops from 99.9% in normal cases to 54.3%, with the probability spread across the options.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/19/2026
First Seen
Valid until: 1/17/2027
Sources
Related Entities
Related Claims
UnverifiedWhen users presented a wrong answer as a confident statement and pushed twice, the model persistence rate dropped from 94.4% to 84.0%.76% similarUnverifiedIn a four-choice scenario, even if the top-1 probability reaches 0.64, the normalized confidence may be only 0.26, which is why the plugin, comparing confidence against the 0.8 threshold, can never pass.76% similarUnverifiedWhen the model claims at least 90% confidence, its actual accuracy is only 56%, making it an extremely poorly calibrated model; therefore ReachNT hides the confidence percentage from tenants, and 42% of dangerous reports are routed to human handling before any decision is made.75% similarUnverifiedRanked by question difficulty, persistence rate was 91.4% on the 60 questions that all models answered correctly, but only 73.5% on the 52 questions that at least one model got wrong, indicating that sycophancy tracks uncertainty.74% similarUnverifiedMost general-purpose large models suffer from overconfidence issues, where their output probabilities reflect token distribution preferences more than genuine uncertainty estimates.73% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1069969API
curl https://kongchang.com/api/v1/knowledge/claims/1069969MCP
get_claim(id=1069969)