Unverified50% confidenceBenchmarkExact time
Among the facts that GPT-5 explicitly said it "did not know," 61% were still answered correctly.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
11/5/2026
First Seen
Valid until: 2/3/2027
Sources
Related Entities
Related Claims
UnverifiedOn FinDendro's controlled facts, GPT-5 answered 76% correctly but claimed to know only 38%.78% similarUnverifiedThe author believes that for 99% of people, continuously using GPT-5.5 with the reasoning level set to medium is sufficient.70% similarUnverifiedWith complete information, GPT-6-luna chose 'no particular feeling' for 65% of hands, the most silent among all models; and it only generates a feeling when it receives both hole cards and betting information.69% similarUnverifiedHow much shared quota roughly one fold of GPT-6 Pro consumes has not yet been clarified in this conversion.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1138977API
curl https://kongchang.com/api/v1/knowledge/claims/1138977MCP
get_claim(id=1138977)