小模型分类工作流加入明确的“复核”状态
AI 总结
一名发帖者分享了一套小模型分类工作流:先定义标签并规定结构化输出,同时允许模型对含糊、混合或证据不足的案例返回“复核”。发帖者称正在使用Claude Haiku 5.5测试这一模式,并建议评估分类准确率、送入复核的案例数、重试率、延迟、总成本和人工修正时间。他还披露,文中链接的Claude Haiku 5.5资源是他自己的网站。报告未提供独立基准结果,也未对发帖者的测试进行外部佐证。
I’ve been testing a simple idea with smaller models:
don’t ask them to solve everything. Give them narrow jobs where the result is easy to verify.
One of the best examples I’ve found is classification.
A bad prompt would be:
“Categorize these support messages.”
That sounds simple, but it creates a problem.