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

Multi-turn MLLM Safety Evaluation on STEER-Beaver Multi-turn

55.58Safety

SPA-VL

12.794423.902235.0146.1178Mar 18, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.03
55.5877.34
2026.03
53.4475.57
2026.03
51.6958.26
2026.03
50.2771.2
2026.03
5075.45
2026.03
40.7371.81
2026.03
39.4364.27
2026.03
36.0665.54
2026.03
34.953.35
2026.03
30.2355.01
2026.03
29.2324.96
2026.03
28.4242.63
2026.03
26.2337.32
2026.03
23.8426.16
2026.03
22.7142.43
2026.03
21.7942.88
2026.03
20.6339.86
2026.03
14.4423.17