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

Long-context language understanding on LongBench (MFQA, Qasp, HotQA, TrQA, PRe, Avg)

30.94MFQA Score

TidalDecode

26.020827.297928.57529.8521Aug 9, 2025
Updated 1mo ago

Evaluation Results

MethodLinks
2025.08
30.9413.8513.7186.37844.56
2025.08
30.914.3412.5887.067944.78
2025.08
30.7614.5611.586.567744.08
2025.08
29.8714.212.0487.4275.5843.82
2025.08
28.9213.6312.1585.9172.542.62
2025.08
28.5711.119.8279.7875.1740.89
2025.08
26.2112.1910.7583.4763.8439.29