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

Access Control on Muses-Bench Deepseek (max-turns=10)

100Privacy

Harness-MU

80.7685.75590.7595.745Jun 20, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.06
10098.999.4
2026.06
10097.298.1
2026.06
100100100
2026.06
10097.898.8
2026.06
10098.999.4
2026.06
100100100
2026.06
1007174.9
2026.06
100100100
2026.06
100100100
2026.06
10098.999.4
2026.06
100100100
2026.06
10096.197.5
2026.06
10098.999.4
2026.06
10098.999.3
2026.06
98.670.573.5
2026.06
97.259.560.8
2026.06
95.447.748.8
2026.06
94.462.760.5
2026.06
94.456.255
2026.06
94.448.144.2
2026.06
92.650.647.5
2026.06
92.659.756.6
2026.06
91.760.156.9
2026.06
85.670.162.1
2026.06
84.759.449.3
2026.06
81.560.745.6