AnyJev: typed decisions from open models

A Python library that turns open causal models into Choice, Noul and Score decisions with access to logits or hidden states.

Notes

Its L0 uses no labels, L1 calibrates with 100–500 labels per question, and L2 fits a model- and question-specific head with 100–300 labels; the repository marks its comparisons as author-published.

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

Destinations

Filed under

• #local
• #replica
• #python
• #calibration

In collections

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