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

Finance

Benchmarks

Task NameDataset NameSOTA ResultTrend
Language ModelingFinance (Fin)
PPL Change (%)0
28
Machine Translation (Zh to En)Finance
SacreBLEU49.1
26
AI-generated text detectionFinance
AUC0.998
24
Machine Translation (En to Zh)Finance
SacreBLEU39.6
20
Time Series ClassificationFinance (test)
F1 Score63.1
19
Named Entity RecognitionFinance (test)
F1 Score87.25
14
Machine-generated text detectionFinance Llama-3-70B-Instruct (test)
AUC0.995
12
AI-generated text detectionFinance GPT-3.5 Turbo
AUC98.7
12
Data ExtractionFinance D2
Match Ratio (Mean)47.6
11
Financial ReasoningFinance
Accuracy52.45
11
Sentiment ClassificationFinance (test)
Accuracy90.8
11
Sentiment ClassificationFinance
F1 Score89.41
11
Open-ended GenerationFinance
ROUGE-Lsum29.19
8
Text ClassificationFinance
Label Quality62.3
5
Partition SelectionFinance
Output Size17,695
4
Theme Label QualityFinance Out-of-domain
ROUGE-142.4
4
Theme DistributionFinance Out-of-domain
Accuracy55.8
4
Cross-domain generalizationFinance (test)
Accuracy99.3
3
Parametric Dynamical System ModelingFinance Extrapolation
TtT0.225.9
2
Parametric Dynamical System ModelingFinance Interpolation
TtT0.245
2
Sentiment AnalysisFinance
Accuracy66
2
Showing 21 of 21 rows