Python — Local reproduction: implements Jev's typed `Choice`/`Score`/`Noul` interface on commodity zero-shot NLI models and makes the confidence honest with temperature scaling and conformal abstention, shipping a reproducible calibration eval (ECE 0.170 to 0.071, cross-validated) that runs offline with no API key.

rupeshpoojary9evaluation-benchmarking2 stars · 0 forks

First discovered 2026-09-20, last refreshed 2026-09-20. Descriptions and stats are pulled from the project's own GitHub repo and refreshed automatically — they aren't independently verified by JevTracks beyond the initial eligibility check.