by Heman10x-NGU · Agent Tool · ★ 295
openJev-verdict-2.0: Non-Autoregressive System 1 Decision Engine -brightgreen) -success) openJev-verdict-2.0 is an open-source, non-autoregressive foundational decision model engineered for deterministic software automation, inspired by TypeSafe AI's Jev and Reinforcement Learning for Calibrated Decisions (RLCD). While standard Large Language Models waste compute generating text strings that deterministic software must parse and validate, openJev-verdict-2.0 evaluates typed decision schemas (, , ) in non-autoregressive forward passes (20–25 ms per decision) with actuarial-grade c
| Stars | 295 |
| Forks | 47 |
| Language | Python |
| Category | Agent Tool |
| Quality Score | 59.9815173442794/100 |
| Open Issues | 3 |
| Last Updated | 2026-09-20 |
| Created | 2026-09-19 |
| Platforms | browser, python |
| Est. Tokens | ~15k |
These tools work well together with openJev-verdict-2.0 for enhanced workflows:
Explore other popular agent tool tools:
openJev-verdict-2.0 is Calibrated 151M Non-Autoregressive Decision Engine beating TypeSafe Jev & Laya on LocalLLaMA/typed-decisions (77.10% acc, 0.0636 Brier, 0.0144 ECE). It is categorized as a Agent Tool with 295 GitHub stars.
openJev-verdict-2.0 is primarily written in Python. It covers topics such as agentic-ai, brier-score, calibration.
You can find installation instructions and usage details in the openJev-verdict-2.0 GitHub repository at github.com/Heman10x-NGU/openJev-verdict-2.0. The project has 295 stars and 47 forks, indicating an active community.