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

MemoTrap

Benchmarks

Task NameDataset NameSOTA ResultTrend
completion taskMemoTrap (test)
Accuracy61.83
30
Memorization Trap DetectionMemoTrap
Macro Score71.96
15
Knowledge RetrievalMemoTrap
Proverb Score42.5
12
Hallucination PredictionMemoTrap
Accuracy65
6
Knowledge Conflict ResolutionMemoTrap
Micro Accuracy77.35
4
Showing 5 of 5 rows