Unverified50% confidenceFactExact time

promptfoo does not raise an error when the sum of scoring rubric weights is not 1.0; instead it computes a weighted average using the given weights.

1

Sources

50%

Confidence

Long-term

Relevance

10/13/2026

First Seen

Sources

LLM Judge Rubrics in Practice: promptfoo vs DeepEval

Related Entities

promptfooproduct

Related Claims

Unverified拦截回归数作为绝对计数指标不需要解释0到1的评分标尺，对董事会更有说服力71% similar Unverified永远不要用单次调用结果做LLM质量判断，应重复运行并计算通过率64% similar Unverified应通过重复运行计算通过率（pass rate）而非依赖单次结果来判定LLM质量63% similar Unverified流式响应的正确计费实现应只在流结束时根据最终usage字段一次性计费，或只累加每个chunk的增量Token数62% similar

Cite This Claim

Stable URI

https://kongchang.com/claim/1018293

API

curl https://kongchang.com/api/v1/knowledge/claims/1018293

MCP

get_claim(id=1018293)