Jeeves: Reasoning Improves Jev-Like Decision Models

A developer released Jeeves, a Jev-like decision model built on Qwen3.5-9B with LoRA, a pointer head, CISPO training, and a block-4 diffusion drafter. It scores 0.889 on unseen test data versus 0.822 for Kev-9B and 0.857 for Jev, and 0.935 on JevBench's public hard tier; latency on one H100 is about 0.3s without thinking and 3.3s median with it, with code, weights, and data open-sourced.

• JevBench: a reproducible benchmark for typed decision models2026-09-22
• Home-Trained 0.8B Decision Model Jeff Runs in ~30ms2026-09-28
• Turning GLM-5.3-Flash into a Jev-like decision model2026-09-27
• Developer has Jev play Pokémon Red, pushing AI to a more complex game2026-09-25
• Open-source Kev 4B tested side by side with Jev2026-09-25