Unverified50% confidenceBenchmarkExact time
In actual testing, the draft acceptance rate was only approximately 41%, with the draft model outputting 1,055 tokens but only approximately 438 being accepted.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/6/2026
First Seen
Valid until: 12/5/2026
Sources
Related Entities
Related Claims
Unverified在一个396个文件的大型混合职责项目上测试时，召回率几乎跌到零65% similarUnverifiedFactoryDroid running the same Opus model submitted a score of 69.9%, while CloudCode scored 58%, but this data was self-reported by the vendors.62% similarUnverified2024 年的多项基准测试表明，经过严格安全对齐的模型在代码生成任务上的拒绝率比基座模型高出 15-30%，在创意写作任务上的自我审查率高达 40%61% similarUnverifiedIncluding specific evaluation dimensions in prompts (such as format, word count, key points to cover, and items to exclude) can improve model output quality by an average of 30%–50% compared to vague prompts.61% similarUnverifiedIn a test-paper grading case, Cowork generated an analysis and HTML report containing an average score of approximately 72%, a minimum of 30 points, and a maximum of 100 points through a single instruction.61% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/863786API
curl https://kongchang.com/api/v1/knowledge/claims/863786MCP
get_claim(id=863786)