Automated README summary
Automatically summarized from the README on 3 Nov 2026; not a run by jevbooks. Run instructions and author-reported outcomes still need verification.
What it is
A Python benchmark harness that runs Jev-style System One decision models locally on Apple Silicon Macs and compares them on English and Korean customer tickets, BANKING77 intent classification, transfer-v4, typed-decisions, NSMC and KLUE-YNAT.
How it uses Jev
Each engine is served behind a /v1/systemone endpoint that accepts typed questions and returns calibrated probabilities; bench/run.py sends the same question sets to every engine and scores accuracy, latency and order-flip, with paired McNemar tests separating engines on some sets.
Primitives:choicescorenoul
Technique worth stealing
Serve every model behind one /v1/systemone contract so heterogeneous engines can be scored on identical questions.
Try it
brew install vllm-metal, install Ollaya, run uv sync, then uv run python bench/run.py --smoke --engine clm.
Model judgments from README excerpts
Project signals use repository metadata, an existing summary and the first 700 README characters, not the full project. “Not found” means not detected in that material; it does not establish absence. These are the original judgments with recorded models and dates, not source audits or run results.
Models & evaluation dates
- Inclusion gate
jev-1.14.0- Project signals
jev-1.14.0- Category
jev-1.14.0- Pattern tags
jev-1.14.0
Evidence
These signals use repository metadata, an existing summary and at most 700 cleaned README characters. ≥ 0.60 means reported in that excerpt; ≤ 0.40 means not found there. The middle band is inconclusive. None is a source-code audit or a test result. The runnable-tool question excludes research experiments; the recommendation question asks for a learning shortlist, not a quality verdict.
- Jev is core to the projectFound in material0.87
- Shows a System One patternnot found in the excerpt0.29
- Describes handling uncertaintynot found in the excerpt0.03
- Reports measured numbersnot found in the excerpt0.17
- Is a runnable toolnot found in the excerpt0.19
- Top-10 learning shortlist signalnot found in the excerpt0.24
- Is a replica of the modelFound in material0.84
- Solves a common problem (0–2)score on a 0–2 scale1.01 / 2
- About JevFound in material0.99
no-judgment band (0.40–0.60)
no-judgment band (0.40–0.60)
The Jev model and date for each judgment stage are listed above. Card written by DeepSeek V4.1 Flash from the README on 3 Nov 2026.