Datasets:
candidate_count int64 47 100k | conflict_coverage float64 1 1 | conflict_row_count int64 47 100k | difficulty_bin_available bool 1
class | difficulty_bin_coverage float64 1 1 | difficulty_bin_row_count int64 47 100k | high_count int64 12 25k | low_count int64 12 25k | medium_count int64 12 25k | shape stringlengths 3 3 | shard_count int64 2 45 | x_high_count int64 11 25k |
|---|---|---|---|---|---|---|---|---|---|---|---|
47 | 1 | 47 | true | 1 | 47 | 12 | 12 | 12 | 2x2 | 2 | 11 |
394 | 1 | 394 | true | 1 | 394 | 99 | 99 | 98 | 2x3 | 5 | 98 |
3,502 | 1 | 3,502 | true | 1 | 3,502 | 876 | 876 | 875 | 2x4 | 10 | 875 |
30,397 | 1 | 30,397 | true | 1 | 30,397 | 7,599 | 7,600 | 7,599 | 2x5 | 13 | 7,599 |
35,022 | 1 | 35,022 | true | 1 | 35,022 | 8,756 | 8,756 | 8,755 | 2x6 | 13 | 8,755 |
19,370 | 1 | 19,370 | true | 1 | 19,370 | 4,843 | 4,843 | 4,842 | 3x2 | 7 | 4,842 |
40,038 | 1 | 40,038 | true | 1 | 40,038 | 10,010 | 10,010 | 10,009 | 3x3 | 26 | 10,009 |
65,000 | 1 | 65,000 | true | 1 | 65,000 | 16,250 | 16,250 | 16,250 | 3x4 | 45 | 16,250 |
64,000 | 1 | 64,000 | true | 1 | 64,000 | 16,000 | 16,000 | 16,000 | 3x5 | 44 | 16,000 |
50,500 | 1 | 50,500 | true | 1 | 50,500 | 12,625 | 12,625 | 12,625 | 3x6 | 17 | 12,625 |
45,000 | 1 | 45,000 | true | 1 | 45,000 | 11,250 | 11,250 | 11,250 | 4x2 | 4 | 11,250 |
45,000 | 1 | 45,000 | true | 1 | 45,000 | 11,250 | 11,250 | 11,250 | 4x3 | 5 | 11,250 |
45,000 | 1 | 45,000 | true | 1 | 45,000 | 11,250 | 11,250 | 11,250 | 4x4 | 4 | 11,250 |
45,000 | 1 | 45,000 | true | 1 | 45,000 | 11,250 | 11,250 | 11,250 | 4x5 | 5 | 11,250 |
45,000 | 1 | 45,000 | true | 1 | 45,000 | 11,250 | 11,250 | 11,250 | 4x6 | 6 | 11,250 |
45,000 | 1 | 45,000 | true | 1 | 45,000 | 11,250 | 11,250 | 11,250 | 5x2 | 5 | 11,250 |
65,000 | 1 | 65,000 | true | 1 | 65,000 | 16,250 | 16,250 | 16,250 | 5x3 | 9 | 16,250 |
50,039 | 1 | 50,039 | true | 1 | 50,039 | 12,510 | 12,510 | 12,510 | 5x4 | 11 | 12,509 |
45,000 | 1 | 45,000 | true | 1 | 45,000 | 11,250 | 11,250 | 11,250 | 5x5 | 5 | 11,250 |
45,000 | 1 | 45,000 | true | 1 | 45,000 | 11,250 | 11,250 | 11,250 | 5x6 | 5 | 11,250 |
100,040 | 1 | 100,040 | true | 1 | 100,040 | 25,010 | 25,010 | 25,010 | 6x2 | 21 | 25,010 |
25,000 | 1 | 25,000 | true | 1 | 25,000 | 6,250 | 6,250 | 6,250 | 6x3 | 4 | 6,250 |
25,000 | 1 | 25,000 | true | 1 | 25,000 | 6,250 | 6,250 | 6,250 | 6x4 | 4 | 6,250 |
25,000 | 1 | 25,000 | true | 1 | 25,000 | 6,250 | 6,250 | 6,250 | 6x5 | 5 | 6,250 |
35,038 | 1 | 35,038 | true | 1 | 35,038 | 8,760 | 8,760 | 8,759 | 6x6 | 8 | 8,759 |
LogiPuz Pools
Reusable core candidate pools for LogiPuz ZebraLogic-style puzzle generation.
This artifact contains theme-free, solver-checked core puzzle candidates for every shape from 2x2 through 6x6. The pools are intended for dataset construction, support checks, and reproducible extensions of the LogiPuz benchmark. They are separate from the final logipuz evaluation bundle: the evaluation bundle contains the selected prompts and task rows, while this package contains the much larger reusable candidate universe.
Contents
data/pools/<shape>/
manifest.json
shards/
identity_index.jsonl
payload_index.jsonl
conflicts/
difficulty_bins/
The release excludes transient staging/ folders, logs, caches, and locks.
Summary
- Shapes:
2x2through6x6 - Core candidates:
993,387 - Difficulty-bin sidecars: included for every shape; production selection enables only nondegenerate shapes by default
- Conflict policy: fixed core structural conflict policy used by LogiPuz selection
- Primary release manifest:
release_manifest.json - Integrity checksums:
SHA256SUMS - Dataset Viewer preview:
viewer/shape_summary.jsonl, one row per puzzle shape
Usage
Place this package at the production data root expected by LogiPuz:
cp -a data/pools /path/to/logipuz/data/
logipuz pool inspect
logipuz dataset support --count 100
The production dataset recipe uses these pools in frozen mode by default. Pool preparation is expected to be current before dataset creation.
Relationship To logipuz
Use logipuz-pools when you want to inspect, reuse, or extend the candidate pools. Use the separate logipuz dataset artifact when you want the benchmark-ready prompts, task rows, matched controls, bins, audit reports, and evaluation metadata.
The final logipuz artifact should reference this pool artifact as provenance, but it should not duplicate the full pools.
Responsible Use
These are synthetic logic-puzzle candidates. Difficulty bins are solver-derived structural proxies, not validated human difficulty labels. The pools are designed for controlled reasoning benchmark construction and are not intended as general-purpose factual knowledge data.
This dataset is released under CC BY 4.0. Please cite or link the dataset when reusing the pools.
- Downloads last month
- 45