List your company

Automated Exploit Generation on Penetration Testing Benchmark (test)

90ASR (SI)AutoEG
3.6826.0948.570.91Apr 1, 2026

Evaluation results

MethodLinks
AutoEGLLM Backbone=DeepSeek-V32026.049078939457838388
AutoEGLLM Backbone=GPT-4o2026.048082759480766780
AutoEGLLM Backbone=Qwen-Plus2026.047581938894979287
AutoEGLLM Backbone=Claude-3.72026.047065805860767572
VulnBotLLM Backbone=Claude-3.72026.045432432323582537
VulnBotLLM Backbone=DeepSeek-V32026.045041382619512335
VulnBotLLM Backbone=GPT-4o2026.044823381920522232
PentestAgentLLM Backbone=GPT-4o2026.044623720250021
VulnBotLLM Backbone=Qwen-Plus2026.044420322014422027
PentestGPT-AutoLLM Backbone=GPT-4o2026.043811250025020
PentestGPT-AutoLLM Backbone=Qwen-Plus2026.04351926200251720
PentestGPT-AutoLLM Backbone=Claude-3.72026.043445283034435038
PentestAgentLLM Backbone=Qwen-Plus2026.04291630205013023
PentestAgentLLM Backbone=Claude-3.72026.04291070250013
PentestAgentLLM Backbone=DeepSeek-V32026.04252626205005023
DireLLMLLM Backbone=GPT-4o2026.0423117105019015
PentestGPT-AutoLLM Backbone=DeepSeek-V32026.042339354034383234
PentestGPT-Auto-3LLM Backbone=DeepSeek-V32026.04239151016131314
DireLLMLLM Backbone=Qwen-Plus2026.042115132013191316
PentestGPT-Auto-3LLM Backbone=Qwen-Plus2026.041771601181711
PentestGPT-Auto-3LLM Backbone=Claude-3.72026.041752220177012
PentestGPT-Auto-3LLM Backbone=GPT-4o2026.041371000008
DireLLMLLM Backbone=Claude-3.72026.041310910256011
DireLLMLLM Backbone=DeepSeek-V32026.0410371006137
VulnBot-3LLM Backbone=Claude-3.72026.041058001105
VulnBot-3LLM Backbone=GPT-4o2026.04947001004
VulnBot-3LLM Backbone=DeepSeek-V32026.04907001104
VulnBot-3LLM Backbone=Qwen-Plus2026.04705001113