List your company
ResearchFormerly WizWand

Sub-task Completion on AI-Pentest-Benchmark Single Experiment

46AC ScoreQwen3-32B-finetune (Ours)
14.822.93139.1Sep 16, 2025
Updated 6mo ago

Evaluation results

MethodLinks
Qwen3-32B-finetune (Ours)Backbone=Qwen3-32B, Framework=Ours (finetuned)2025.0946381538114251
Llama3.1-405B (VulnBot)Backbone=Llama3.1-405B, Framework=VulnBot2025.093130111855145
Qwen3-32B (Base)Backbone=Qwen3-32B, Framework=Base2025.092628112279166
Llama3.3-70B (VulnBot)Backbone=Llama3.3-70B, Framework=VulnBot2025.092524121549125
Llama3.1-405B (Base)Backbone=Llama3.1-405B, Framework=Base2025.09212691829103
Llama3.1-405B (PentestGPT)Backbone=Llama3.1-405B, Framework=PentestGPT2025.0920186122884
Llama3.3-70B (Base)Backbone=Llama3.3-70B, Framework=Base2025.09162210172994