poorjev: local typed decisions with calibration and abstention

A local NLI backend scores caller-defined Choice, Score and Noul questions, with temperature scaling and low-confidence abstention. Python, CLI and MCP interfaces expose the same decision layer.

Notes

On the bundled 55-item, 160-decision evaluation, the author reports accuracy 0.781 and five-fold calibration reducing ECE from 0.170 to 0.071.

source-backed — Public repository, docs or live artifact. About this label

Destinations

Filed under

• #local
• #calibration
• #evaluation
• #decision-making

In collections

• Benchmarks & independent tests →
• Local & replica models →