タグ：intent-classification

Pre-registered independent eval of TypeSafe Jev against a nano-class LLM, a frontier LLM, and a supervised encoder (Banking77 + CLINC150 zero-shot)

Curated list of tools related to system one models

Finite-sample guarantees for Jev (TypeSafe's System One). Conformal risk control turns calibrated probabilities into certified routing thresholds; prediction-powered inference audits them. 2,412 decisions on CLINC150 for $0.23 — including the shift and prevalence cases where the guarantee breaks.

Calling TypeSafe's Jev from Java with a hand-typed API contract. Small app to demo triage, zero-shot classification of commands from MASSIVE's 60 intents. And live batch classification with confidence-based routing.