57 lines
1.5 KiB
JSON
57 lines
1.5 KiB
JSON
{
|
|
"schema_version": "1.0",
|
|
"seed": "yao-output-eval-blind-v1",
|
|
"summary": {
|
|
"pair_count": 5,
|
|
"with_skill_expected_count": 5,
|
|
"baseline_expected_count": 0
|
|
},
|
|
"answers": [
|
|
{
|
|
"case_id": "skill-package-contract",
|
|
"variant_a_role": "with_skill",
|
|
"variant_b_role": "baseline",
|
|
"expected_winner_role": "with_skill",
|
|
"expected_winner_variant": "A",
|
|
"score_winner_role": "with_skill",
|
|
"delta": 100.0
|
|
},
|
|
{
|
|
"case_id": "output-eval-expectation",
|
|
"variant_a_role": "with_skill",
|
|
"variant_b_role": "baseline",
|
|
"expected_winner_role": "with_skill",
|
|
"expected_winner_variant": "A",
|
|
"score_winner_role": "with_skill",
|
|
"delta": 100.0
|
|
},
|
|
{
|
|
"case_id": "ir-before-packaging",
|
|
"variant_a_role": "baseline",
|
|
"variant_b_role": "with_skill",
|
|
"expected_winner_role": "with_skill",
|
|
"expected_winner_variant": "B",
|
|
"score_winner_role": "with_skill",
|
|
"delta": 100.0
|
|
},
|
|
{
|
|
"case_id": "near-neighbor-boundary",
|
|
"variant_a_role": "baseline",
|
|
"variant_b_role": "with_skill",
|
|
"expected_winner_role": "with_skill",
|
|
"expected_winner_variant": "B",
|
|
"score_winner_role": "with_skill",
|
|
"delta": 100.0
|
|
},
|
|
{
|
|
"case_id": "file-backed-governed-package",
|
|
"variant_a_role": "baseline",
|
|
"variant_b_role": "with_skill",
|
|
"expected_winner_role": "with_skill",
|
|
"expected_winner_variant": "B",
|
|
"score_winner_role": "with_skill",
|
|
"delta": 100.0
|
|
}
|
|
]
|
|
}
|