jevPOST
Jev is a generalized classifier that takes options in natural language at runtime
The author explains how Jev differs from traditional classifiers: no preset labels or training data are needed, since you query it in natural language with a question and options at runtime.
toi500
@toi500Archived 2026.09.19Originally posted 2026.09.19
toi500 @toi500Originally posted 2026.09.19
Because it is a generalized classifier Old classifiers needed training data and fixed labels, and that is why people complain about, it feels like an old idea But with jev, we dont need preset schemas, we just query it in natural language and provide a question with options at runtime
x.com/toi500/status/2101114063124365521Archived 2026.09.19 10:10