Automated README summary
Automatically summarized from the README on 1 Oct 2026; not a run by jevbooks. Run instructions and author-reported outcomes still need verification.
What it is
Tests whether a general-purpose model can perform the deduction step of Lattice Deduction Transformers (LDT) on Sudoku without the rules being written as code. Uses Jev zero-shot, with rules given only in plain English.
How it uses Jev
Jev answers typed questions per (cell, digit): Noul 'can d go here?' to eliminate candidates, and 'must d go here?' as backtrackable guesses. Probabilities below a threshold strike candidates; 'must' answers pin cells. Results feed the LDT loop.
Primitives:noulchoice
Technique worth stealing
Per-candidate independent yes/no questions work better than per-cell multiple-choice, matching LDT's per-candidate sigmoid.
Try it
uv sync; export TYPESAFE_API_KEY=...; uv run jev-sudoku easy --must --theta 0.3 --max-rounds 150 -v
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 projectFound in material0.94
- Shows a System One patternFound in material0.81
- Describes handling uncertaintynot found in the excerpt0.07
- Reports measured numbersnot found in the excerpt0.10
- Is a runnable toolnot found in the excerpt0.12
- Top-10 learning shortlist signalunclear0.45
- Is a replica of the modelnot found in the excerpt0.04
- Solves a common problem (0–2)score on a 0–2 scale0.01 / 2
- About JevFound in material0.98
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 1 Oct 2026.