Datasets:

csoai
/

compbench

Modalities:

Tabular

Text

Size:

< 1K

License:

Dataset card Data Studio Files Files and versions

xet

Community

Dataset Viewer

Auto-converted to Parquet Duplicate

Split (1)

train · 2 rows

timestamp stringclasses 2 values	model stringclasses 2 values	backend stringclasses 1 value	prompt_mode stringclasses 1 value	categories dict	composite float64 78.2 78.2	correct int64 86 86	total int64 110 110
2026-08-02T22:56:06.501322+00:00	sov-sovereign-v4:latest	ollama-local	neutral	{ "agentic": { "correct": 15, "total": 15, "pct": 100 }, "coding": { "correct": 15, "total": 15, "pct": 100 }, "general": { "correct": 15, "total": 15, "pct": 100 }, "governance": { "correct": 6, "total": 15, "pct": 40 }, "instruction": { "correct": ...	78.2	86	110
2026-08-02T22:49:12.469889+00:00	sov33-v7:latest	ollama-local	neutral	{ "agentic": { "correct": 15, "total": 15, "pct": 100 }, "coding": { "correct": 15, "total": 15, "pct": 100 }, "general": { "correct": 15, "total": 15, "pct": 100 }, "governance": { "correct": 6, "total": 15, "pct": 40 }, "instruction": { "correct": ...	78.2	86	110

compbench

CompBench — compliance cross-walk measurement runs (neutral prompt set). Raw measured JSON per model.

Measurement, not certification. All scores are measured runs on published harnesses; score bands are descriptive, not accreditation. UNMEASURED is reported, never scored as fail.

Downloads last month: 27

Number of rows:

Total file size:

4.6 kB