| family,dataset,task,exposure,n,ours,jev,dataset_url,source,reference | |
| heldout_intent_clinc,CLINC150,Intent routing,held-out source,300,95.0,94.33333333333333,https://huggingface.co/datasets/clinc/clinc_oos,reports/selfjev_4b_treeserver/test/report.json,reports/external/full/typesafe_jev-latest/report.json | |
| heldout_topic_dbpedia,DBpedia,Topic classification,held-out source,300,98.0,98.0,https://huggingface.co/datasets/fancyzhx/dbpedia_14,reports/selfjev_4b_treeserver/test/report.json,reports/external/full/typesafe_jev-latest/report.json | |
| heldout_question_type_trec,TREC,Question type,held-out source,300,91.33333333333333,94.0,https://huggingface.co/datasets/SetFit/TREC-QC,reports/selfjev_4b_treeserver/test/report.json,reports/external/full/typesafe_jev-latest/report.json | |
| heldout_emotion_multiclass,Emotion,Emotion classification,held-out source,300,57.333333333333336,57.666666666666664,https://huggingface.co/datasets/dair-ai/emotion,reports/selfjev_4b_treeserver/test/report.json,reports/external/full/typesafe_jev-latest/report.json | |
| heldout_boolq,BoolQ,Reading comprehension,held-out source,300,87.66666666666667,90.66666666666667,https://huggingface.co/datasets/google/boolq,reports/selfjev_4b_treeserver/test/report.json,reports/external/full/typesafe_jev-latest/report.json | |
| heldout_sentiment_sst2,SST-2,Sentiment,held-out source,300,91.33333333333333,96.66666666666667,https://huggingface.co/datasets/stanfordnlp/sst2,reports/selfjev_4b_treeserver/test/report.json,reports/external/full/typesafe_jev-latest/report.json | |
| hf_intent_banking77,Banking77,Banking intent,seen source,300,96.33333333333333,95.66666666666667,https://huggingface.co/datasets/mteb/banking77,reports/selfjev_4b_treeserver/test/report.json,reports/external/full/typesafe_jev-latest/report.json | |
| hf_topic_agnews,AG News,News topic,seen source,300,90.33333333333333,87.33333333333333,https://huggingface.co/datasets/fancyzhx/ag_news,reports/selfjev_4b_treeserver/test/report.json,reports/external/full/typesafe_jev-latest/report.json | |
| hf_sentiment_tweets,TweetEval,Tweet sentiment,seen source,300,67.33333333333333,64.33333333333333,https://huggingface.co/datasets/cardiffnlp/tweet_eval,reports/selfjev_4b_treeserver/test/report.json,reports/external/full/typesafe_jev-latest/report.json | |
| hf_nli,MNLI,Entailment (binary),seen source,300,95.0,92.66666666666667,https://huggingface.co/datasets/nyu-mll/glue,reports/selfjev_4b_treeserver/test/report.json,reports/external/full/typesafe_jev-latest/report.json | |
| hf_emotions_multilabel,GoEmotions,Emotion tags (exact set),seen source,300,47.666666666666664,31.666666666666668,https://huggingface.co/datasets/google-research-datasets/go_emotions,reports/selfjev_4b_treeserver/test/report.json,reports/external/full/typesafe_jev-latest/report.json | |