Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

CrisisMMD

Benchmarks

Task NameDataset NameSOTA ResultTrend
Multimodal ClassificationCrisisMMD
BC Accuracy72.33
16
Humanitarian classificationCrisisMMD (test)
Macro F183.4
14
Multimodal ClassificationCrisisMMD humanitarian (target domain)
Acc (Pred Deq | Dfl, Dwf, Dhu)75.1
6
Multimodal ClassificationCrisisMMD informative target domain
Accuracy (Dfl, Dwf, Dhu -> Deq)88.6
6
Text Rationale ExtractionCrisisMMD
Token F182.6
4
Rationale Faithfulness EvaluationCrisisMMD v1.0 (test)
Comprehensiveness51.4
3
Showing 6 of 6 rows