ULMFit

Reported on 4 benchmarks across 4 tasks

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Natural Language Processing3 results

Cross-LingualonReddit Ideological and Extreme Bias Dataset
weighted-F1 score
76.8
best: 79.1 (SVM)
Text ClassificationonThreatGram 101 - Extreme Telegram Data
weighted-F1 score
55.7
best: 66.2 (GPT-2)
Cross-Lingual Document ClassificationonReddit Ideological and Extreme Bias Dataset
weighted-F1 score
76.8
best: 79.1 (SVM)

Methodology1 result

ClassificationonThreatGram 101 - Extreme Telegram Data
weighted-F1 score
55.7
best: 66.2 (GPT-2)