haruspex / llm /eval.json
TyrusRC's picture
Upload folder using huggingface_hub
858176a verified
Raw History Blame Contribute Delete
105 Bytes
{
"accuracy": 0.9452,
"ece": 0.0548,
"brier": 0.0548,
"test_n": 73,
"dataset": "llm held-out"
}