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

Long-context Language Modeling on RULER (Detailed Metrics)

100S1 Score

MoE Transformer

99.79299.84699.999.954Jun 2, 2026Jun 6, 2026Jun 11, 2026Jun 16, 2026Jun 21, 2026Jun 26, 2026Jul 1, 2026
Updated 23d ago

Evaluation Results

MethodLinks
2026.06
10010073.87227.6636.540.813.326.823.947.3--
2026.06
10099.881.874.245.438.437.440.515.116.38.150.6--
2026.07
100100100100989610098-88.6799.291.565662.83
2026.07
10010010098988810098.5-90.6796.490.425461.5
2026.07
10010010098989297.599.5-8696.489.855258.83
2026.07
100100100100100100100100-9299.292.957054.17
2026.07
1001001001001009898.599-8899.292.136854.83
2026.07
10010098100981009998-89.391.290.475854.17
2026.07
1001001009898989698.5-8410090.055652.11
2026.07
10010010098989896.5100-76.6710089.335450.83
2026.07
1001001009098949096.5-76.6710086.945048.11
2026.06
99.810083.8704.821.239.43226.443.39.348.2--
2026.06
99.81009366.212.811.848.250.431.126.418.450.7--