Does decomposing one question into twelve help?

A thread testing whether many narrow Jev questions beat one direct question on a classification task with trap phrasing.

Notes

Single question scored 100%, the 12-dimension version 98%; 25,174 API calls for $1.43.

author demo — Result shown by its author. About this label

Destinations

X post only — no separate service, repository or docs URL.

Filed under

• #evaluation
• #decomposition

In collections

• Benchmarks & independent tests →