OpenJev (Verdict)

# OpenJev (Verdict): Non-Autoregressive Decision Engine (151M)

[](https://huggingface.co/heman10x/rlcd-modernbert-151m)
[](https://github.com/Heman10x-NGU/Verdict-open-jev)
[](#running-the-in-browser-webgpu-playground)
[](#upstream-credits-and-licenses)
[](#reinforcement-learning-for-calibrated-decisions-rlcd)
[](LICENSE)

**OpenJev (Verdict)** is an open-source, post-trained foundational decision model designed for structured software workflows, inspired by **TypeSafe AI\'s Jev** and **Reinforcement Learning for Calibrated Decisions (RLCD)**. 

Instead of generating free-form conversational text that software must parse and validate, OpenJev accepts unstructured input context and evaluates multiple typed questions in a **single non-autoregressive forward pass**. It returns discrete choices, ordinal scores, and binary probabilities with mathematically calibrated confidence values in under 35 milliseconds.

---

## What changed in the inference engine

These are inference fixes, not a retrain. The weights are byte-identical to the published checkpoint. Measured on the 231 public JevBench tasks.

<p align="center">
  <img src="assets/v1.4/benchmark-leaderboard-chart.png" alt="JevBench Score Leaderboard: openJev Verdict v1.4" width="100%">
</p>

<p align="center">
  <img src="assets/v1.4/benchmark-leaderboard-table.png" alt="JevBench Detailed Evaluation Table: openJev Verdict v1.4" width="100%">
</p>

The update addresses three defects in the inference engine:

1. Calibrator auto-loading and removal of the 5-option scope restriction: The engine previously failed to load `calibrator.json` during standalone instantiation, running at uncalibrated temperature 1.0. A scope check also limited calibration exclusively to 5-candidate queries, leaving other cardinalities unscaled.