squaredcuber/loss-aware-dro-repro-artifacts / bundle /v0 /configs /regression_models_table3.json
squaredcuber's picture
download
raw
973 Bytes
{
"schema_version": 1,
"paper_anchor": "Appendix E.2, Table 3",
"x_distribution": {"family": "uniform", "low": -10.0, "high": 10.0},
"noise_distribution": "normal_zero_mean",
"models": [
{"id": 1, "weight": -6.7805, "noise_variance": 564.285},
{"id": 2, "weight": -5.8464, "noise_variance": 625.412},
{"id": 3, "weight": -2.7811, "noise_variance": 699.653},
{"id": 4, "weight": -1.3851, "noise_variance": 710.190},
{"id": 5, "weight": -0.0144, "noise_variance": 783.458},
{"id": 6, "weight": 4.3483, "noise_variance": 846.372},
{"id": 7, "weight": 6.3163, "noise_variance": 915.492},
{"id": 8, "weight": 7.1061, "noise_variance": 922.537},
{"id": 9, "weight": 8.5174, "noise_variance": 932.399},
{"id": 10, "weight": 8.9350, "noise_variance": 978.001}
],
"precision_note": "Values are transcribed from the rounded values printed in Table 3; underlying full-precision author draws are unavailable."
}

Xet Storage Details

Size:
973 Bytes
·
Xet hash:
599615ddba9f4ab32ecd677704a35b17a92c43c26d39581b2ece4cb1d403d0b1

Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.