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

Multi-agent Question Answering on ARC Easy (first 300 questions)

42.33Average Accuracy

TextFullT

27.083631.04183538.9582May 25, 2026
Updated 8d ago

Evaluation Results

MethodLinks
2026.05
42.3342.67505.035.47
2026.05
3939.3372.011.47
2026.05
38.8938.67570.7
2026.05
37.3336.67960.2610.04
2026.05
35.8935.331,717.618.18
2026.05
35.67361,137.8911.58
2026.05
35.3337.671,247.513.15
2026.05
34.7836.671,316.214.48
2026.05
30.6730.67113.591.25
2026.05
27.6727.89324.433.61