Dataset Viewer
Auto-converted to Parquet Duplicate
declared
dict
environment
dict
generated_by
string
minimum_detectable_effect
dict
non_claims
list
paired_report
dict
power_curve
dict
required_seeds
int64
schema
string
{ "bootstrap_seed": 707, "effect": 0.25, "generator_seed": 20260914, "n_seeds": 12, "resamples": 100000, "sd": 0.5 }
{ "numpy": "2.5.3", "python": "3.13.14" }
scripts/evaluation_report.py
{ "alpha": 0.05, "kind": "minimum-detectable-effect", "mde": 0.4043739949587996, "n": 12, "power": 0.8, "schema": "ffb-evaluation-v1", "sd": 0.5, "standard_error": 0.14433756729740646, "z_alpha": 1.9599639845400536, "z_power": 0.8416212335729144 }
[ "This is a synthetic demo of the evaluation-hygiene tooling only; it is not evidence about the fly.", "No biological advantage, general memory, AGI or programming-expertise claim is made.", "A null without a power statement is under-powered, not evidence of absence." ]
{ "ci_high": 0.16497162335117949, "ci_level": 0.95, "ci_low": -0.26030449962310037, "effect": -0.04766635105033473, "holm_p": null, "includes_zero": true, "kind": "paired-report", "mean": -0.04766635105033473, "n": 12, "name": "paired-difference", "per_seed_differences": [ -0.6727754676324071,...
{ "alpha": 0.05, "curve": [ { "mde": 0.9905099528998361, "n": 2, "power_at_reference": 0.10895461755061564, "reference_effect": 0.25 }, { "mde": 0.700396304528242, "n": 4, "power_at_reference": 0.17007504575308763, "reference_effect": 0.25 }, { ...
32
ffb-evaluation-demo-v1

No dataset card yet

Downloads last month
152