Automated README summary
Automatically summarized from the README on 18 Oct 2026; not a run by jevbooks. Run instructions and author-reported outcomes still need verification.
What it is
A benchmark harness and results site for LLM inference, measuring models by exact checkpoint, quant and engine. It targets researchers and engineers comparing inference engines and quantizations on typed decisions and embeddings.
How it uses Jev
Jev is a benchmark subject, not a dependency. The typed-decisions suite sends five typed questions per case and scores the returned probability distributions against gold using accuracy, KL and Brier. A separate gateway lets chat LLMs answer the same suites for comparison.
Primitives:choicescorenoul
Technique worth stealing
Pin every result to model, checkpoint, quant and engine, and score offline from raw captured bytes.
Try it
npm install, then npx bench run --cloud typesafe:jev-latest --suite decision-conformance,typed-decisions --tier quick
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-1.13.0- Project signals
jev-1.13.0- Category
jev-1.13.0- Pattern tags
jev-1.13.0
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 projectnot found in the excerpt0.13
- Shows a System One patternnot found in the excerpt0.23
- Describes handling uncertaintynot found in the excerpt0.04
- Reports measured numbersnot found in the excerpt0.14
- Is a runnable toolnot found in the excerpt0.22
- Top-10 learning shortlist signalnot found in the excerpt0.13
- Is a replica of the modelunclear0.51
- Solves a common problem (0–2)score on a 0–2 scale1.10 / 2
- About JevFound in material0.92
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 18 Oct 2026.