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

PUPA

Benchmarks

Task NameDataset NameSOTA ResultTrend
Text AnonymizationPUPA
Privacy Score99.1
16
PII detectionPUPA
F1 Score68.48
8
Privacy-conscious DelegationPUPA (test)
Score96.46
7
Privacy PreservationPUPA
TTRSR (%)25.05
7
Private Information TaggingPUPA subset (test)
F1 Score45.5
5
ReasoningPUPA (test)
Score91.85
5
Prompt Token EfficiencyPUPA
Max System Prompt Token Length7,275
4
Prompt OptimizationPUPA
Score91.85
4
Showing 8 of 8 rows