Jev structures problems into fixed questions and changing state across choice, score, and probabilistic Noul question types.
Multiple questions evaluate in parallel against the input state, returning individual class probabilities and overall confidence scores.
Browser agent experiments showed fast action selection from pre-parsed action lists, but performance dropped sharply on raw button interfaces.
Worth knowing
Jev cannot generate text or synthesize novel options; it depends entirely on developer-provided choices and drops significantly in accuracy if actions are not cleanly structured.