タグ：calibration

Run the new wave of decision models (Laya, Decider, Kev, Jev) side by side on your Mac. Structured input in, calibrated probabilities out, with honest accuracy, calibration, latency and memory numbers.

A semantic interrupt controller: calibrated semantic channels sampled over a live conversation at 5Hz, firing only on a hysteresis-gated edge. The model decides when; your code decides what. TypeSafe Jev + Deepgram Flux, zero dependencies.

Large typed decision model evaluation benchmark and novel calibration standard (`calibration.json`) for any typed decision inference system, covering 275 distinct use-cases and 27,598 individual decisions for Jev-like open typed decision models.

Bench of typed decision models: Jev vs OpenJev vs Laya, small local LLMs and cheap hosted LLMs on the same zero-shot classification tasks, in English and French.