Unverified50% confidenceBenchmarkExact time
When the probe looks at only one prompt at a time (the conventional setting), even after training on 5,249 questions, Ouro's readout rate is only 0.36 and Qwen's only 0.19.
1
Sources
50%
Confidence
Long-term
Relevance
10/21/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedQwen Next scored only 6/108 (5.56%) in testing due to format reliability issues (17 invalid_json responses, with only 6 entries scorable) rather than insufficient comprehension.71% similarUnverifiedOn the SimpleQA benchmark, o4-mini declines to answer only 1% of questions, with an error rate of 75%.68% similarUnverifiedIn EvoMap's benchmark test, sub-agents actually found 373 correct answers, but after the main agent merged the results, only 217 were retained in the final output.67% similarUnverifiedThe target value for the Erdős minimum overlap problem is 0.38080, where lower is better; the value reported by TTT-Discover is 0.38093, and this run's result of 0.38094 is almost identical to the reported value but still has a tiny gap from the target.66% similarUnverifiedMistral Small 4 scored only 0.699 on unit conversion inputs and produced 265 errors that differed from the correct label by exactly a power of ten.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1078904API
curl https://kongchang.com/api/v1/knowledge/claims/1078904MCP
get_claim(id=1078904)