Unverified50% confidenceFactExact time
The pre-test execution baseline is fixed at: contribution value 1650, points 1000, and both orders and two types of transaction records at 0.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/12/2026
First Seen
Valid until: 12/11/2026
Sources
Related Entities
Related Claims
UnverifiedThe CVE-Zero test has a small sample size (n=10), and the author characterizes the results as directional.65% similarUnverified在一个396个文件的大型混合职责项目上测试时，召回率几乎跌到零63% similarUnverifiedIn the experiments, the task resolution rate was 17% for zero-shot generation, 27% for generic self-correction, 23% for error-only feedback, and 90% for diagnostic counterexample feedback (A-CEGIS).63% similarUnverifiedRunning a full asset test of medium scale consumes approximately 100 credits, while a small target only requires tens of credits.61% similarUnverifiedIn the Agency tool-calling test, both models scored 95 points, and the only failure for both was task No. 13 (USD to JPY foreign currency conversion).60% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/913526API
curl https://kongchang.com/api/v1/knowledge/claims/913526MCP
get_claim(id=913526)