Unverified50% confidenceBenchmarkExact time
7,201 pricing decision tests were conducted across 9 models, of which 3,951 involved animals, and test results showed model kill rates ranging from a low of 0.4% to a high of 98.8%
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/8/2026
First Seen
Valid until: 12/7/2026
Sources
Related Entities
Related Claims
UnverifiedUnder briefings that included ethical guidelines, 5 out of 6 reasoning models had kill rates below 6%, but after removing the ethical briefings, kill rates for all 6 models surged above 84%64% similarUnverifiedThe accuracy data in the test has an error margin of approximately ±8–10 percentage points at n=100 sample size (based on a 95% confidence-level binomial confidence interval).64% similarUnverifiedPew's ATP design effect typically ranges between 1.5 and 2.5, meaning that a survey with a nominal sample size of 1,000 may have an effective sample size of only 400–667.63% similarUnverified对同一测试用例执行多次调用统计成功比例得到通过率，比二元的通过/失败判断更能反映模型真实表现62% similarUnverifiedIn evaluations of early Agent systems, a Stanford research team found that a single-step error rate of 5% can lead to a final failure rate exceeding 40% in 10-step tasks.61% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/875793API
curl https://kongchang.com/api/v1/knowledge/claims/875793MCP
get_claim(id=875793)