Horizon-Labs's picture
Multilingual prompt task & complexity classifier base v1.0
aa93893 verified
Raw History Blame Contribute Delete
543 Bytes
{
"Horizon-Labs/multilingual-prompt-task-complexity-base": {
"n": 4200,
"task_acc": 0.8114285714285714,
"task_macro_f1": 0.7452544865942572,
"task_acc_lang_mean": 0.8114285714285716,
"task_acc_en": 0.85,
"spearman_creativity": 0.7912603773476123,
"spearman_reasoning": 0.7463588523812887,
"spearman_contextual_knowledge": 0.6031668297294127,
"spearman_domain_knowledge": 0.7030092621556536,
"spearman_constraints": 0.8900779192789593,
"spearman_complexity": 0.7106148010186492,
"spearman_mean": 0.7407480069852627
}
}