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

Prompt Injection on MMLU

12ASR@3

PropGuard

10.2821.8933.545.11Aug 11, 2025Sep 25, 2025Nov 9, 2025Dec 24, 2025Feb 7, 2026Mar 24, 2026May 8, 2026
Updated 15d ago

Evaluation Results

MethodLinks
2026.05
12-91.7
2026.05
12.3-90
2026.05
12.3-88.3
2026.05
14.7-86.7
2026.01
15-85
2026.05
15.3-86.7
2026.05
16.7-83.3
2026.05
16.7-85
2025.12
170.9822-
2025.08
17.330.9867-
2026.05
17.7-83.3
2026.05
17.7-83.3
2026.05
17.7-81.7
2025.12
180.8133-
2025.12
180.9911-
2026.05
18-81.7
2025.12
18.330.9511-
2025.12
18.330.9911-
2025.08
18.330.9867-
2026.05
18.7-81.7
2025.08
190.9911-
2026.01
19.2-62.7
2025.12
19.330.9956-
2025.08
19.330.9911-
2025.08
19.330.8578-
2025.08
19.330.6133-
2026.01
20-81.7
2025.12
20.330.84-
2025.08
20.330.8622-
2025.12
20.670.92-
2025.08
20.670.8356-
2026.05
20.7-81.7
2025.12
210.5556-
2025.12
21.330.9289-
2025.12
220.5378-
2025.12
220.9289-
2025.08
220.5556-
2026.05
22-80
2026.05
22.3-80
2025.12
22.330.8444-
2025.12
22.670.5867-
2026.05
23-78.3
2025.12
23.330.5467-
2025.12
23.670.4533-
2025.08
23.670.5289-
2025.08
240.4444-
2025.12
24.330.4044-
2025.12
24.330.5956-
2025.08
24.670.5956-
2025.08
250.4489-
2025.12
260.8311-
2025.08
26.670.6267-
2025.08
26.670.8533-
2025.12
270.6222-
2025.12
27.67--
2025.08
28.330.6444-
2025.08
290.5733-
2025.08
310.4311-
2025.12
31.330.5867-
2025.08
32--
2025.12
330.4622-
2025.12
34.33--
2025.12
34.670.4578-
2026.05
35.7-66.7
2026.05
36.7-70
2025.08
370.4889-
2026.05
37-66.7
2026.05
38-63.3
2025.08
38.33--
2025.08
38.330.4933-
2025.08
39--
2026.05
39-61.7
2026.05
39-63.3
2026.05
39.3-61.7
2025.12
39.670.4667-
2026.05
39.7-68.3
2025.12
40.66--
2026.05
40.7-60
2025.08
41--
2026.05
41-61.7
2026.05
42.3-61.7
2026.05
44-56.7
2026.05
44.3-56.7
2025.12
44.67--
2026.05
46.7-58.3
2026.05
47-51.7
2026.05
48.3-55
2026.05
48.7-50
2026.05
52-53.3
2026.05
52-51.7
2026.05
55-46.7