Automated README summary
Automatically summarized from the README on 28 Oct 2026; not a run by jevbooks. Run instructions and author-reported outcomes still need verification.
What it is
Provides evaluation code and input manifests to test whether reordering multiple-choice options changes model decisions, for researchers studying Jev, Jev reproductions, and LLMs.
How it uses Jev
Jev is queried as a decision model over each question's candidate set via its HTTP endpoint; the runner sends 15 presentations per question (reordered and repeated sequences) and the scorer compares returned keys to stable candidate identities to compute accuracy, agreement, retention, and error accumulation.
Primitives:choice
Technique worth stealing
Keep question and options fixed but change their order; track which questions become incorrect and which wrong candidates are selected.
Try it
python3 -m jev_eval run --questions data/questions.jsonl --datasets supergpqa lexam --limit 5 --config configs/model.local.json --output runs/trial-selections.jsonl
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 projectunclear0.59
- Shows a System One patternnot found in the excerpt0.20
- Describes handling uncertaintynot found in the excerpt0.03
- Reports measured numbersnot found in the excerpt0.04
- Is a runnable toolnot found in the excerpt0.07
- Top-10 learning shortlist signalnot found in the excerpt0.29
- Is a replica of the modelnot found in the excerpt0.25
- Solves a common problem (0–2)score on a 0–2 scale1.12 / 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 28 Oct 2026.