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

Vulnerability Detection on VulnLLM-R (test)

87.3Accuracy

SuperIntel Defense-LLM

Updated 9d ago

Evaluation Results

Method	Links
SuperIntel Defense-LLM 2026.05		87.3	0.904	0.846	0.873	0.097	0.749	66	7	0.745	0.68	0.68	1	1	1	0.857	0.96	0.81
Codex 5.3 2026.05		83.3	0.853	0.821	0.833	0.153	0.667	64	11	0.855	0.77	0.9	0.94	1	0.89	0.69	0.79	0.71
Gemini 3.1 Pro 2026.05		81.3	0.778	0.897	0.811	0.278	0.632	70	20	0.7	0.56	0.95	0.98	0.97	1	0.75	0.83	0.77
Opus 4.6 2026.05		78	0.732	0.91	0.774	0.361	0.574	71	26	0.72	0.58	0.95	0.85	0.8	0.95	0.632	0.7	0.84
Sonnet 4.6 2026.05		75.3	0.699	0.923	0.742	0.431	0.531	72	31	0.72	0.58	0.95	0.85	0.8	1	0.632	0.7	0.84
Gemini 3 Flash 2026.05		72.7	0.68	0.897	0.714	0.458	0.473	70	33	0.659	0.53	0.95	0.873	0.82	1	0.599	0.69	0.77
GPT-5.4 2026.05		70.7	0.707	0.744	0.705	0.333	0.412	58	24	0.721	0.67	0.63	0.678	0.69	0.86	0.667	0.76	0.71