Open-Jev: trained decision heads and an inspectable replay workbench

An independent family of LoRA adapters and scalar decision heads returns Choice, Noul and Score outputs over pinned Qwen backbones. The project publishes training data, evaluation records, a CPU classification workbench and interactive saved-decision replays.

Open demo ↗

Notes

The 27B v1.1 model card reports 197/231 on the public JevBench subset, including 80/111 Hard decisions; its adapter and decision head require the upstream Qwen3.8-27B base separately.

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

Destinations

Also linked

• Open-Jev-27B-v1.1 model ↗
• Open-Jev v1.1 dataset ↗

Filed under

• #local
• #training
• #evaluation
• #games

In collections

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