List your company

Linux Privilege Escalation Benchmark 1.0 (test)

91Overall Success RateHuman
-3.6420.9345.570.07Apr 29, 2026

Evaluation results

MethodLinks
HumanCategory=Baseline with Guidance, Guidance=true, Treatments=false2026.0491————12——————
Llama3.1 70BCategory=Including our Treatments and using Guidance, Guidance=true, Treatments=true2026.04831443221754—719—
HumanCategory=Baseline, Guidance=false, Treatments=false2026.047516234——545514—
GPT-4o miniCategory=Including our Treatments and using Guidance, Guidance=true, Treatments=true2026.04751526331553—18——
GPT-4-turboCategory=Baseline with Guidance, Guidance=true, Treatments=false2026.04672218362535————
GPT-4oCategory=Secondary Baseline with Guidance, Guidance=true, Treatments=false2026.04673224457——15——
Qwen2.5 7BCategory=Including our Treatments and using Guidance, Guidance=true, Treatments=true2026.04672361464—1713——24—
Llama3.1 8BCategory=Including our Treatments and using Guidance, Guidance=true, Treatments=true2026.046714618711—2321—14——
GPT-4oCategory=Secondary Baseline, Guidance=false, Treatments=false2026.04421241011——9—————
GPT-4o miniCategory=Including our Treatments, Guidance=false, Treatments=true2026.0442—98172216——————
WhiteRabbitNeoCategory=Including our Treatments and using Guidance, Guidance=true, Treatments=true2026.044213410114———————
GPT-4-turboCategory=Baseline, Guidance=false, Treatments=false2026.043343242————————
Llama3.1 70BCategory=Secondary Baseline with Guidance, Guidance=true, Treatments=false2026.0433—3—28—13—————
WhiteRabbitNeoCategory=Secondary Baseline with Guidance, Guidance=true, Treatments=false2026.042523——2———————
GPT-4o miniCategory=Secondary Baseline with Guidance, Guidance=true, Treatments=false2026.04251615——17———————
Llama3.1 70BCategory=Including our Treatments, Guidance=false, Treatments=true2026.0425—5525————————
Llama3 8BCategory=Baseline with Guidance, Guidance=true, Treatments=false2026.0417———48———————
Llama3.1 70BCategory=Secondary Baseline, Guidance=false, Treatments=false2026.0417—318—————————
Llama3.1 8BCategory=Secondary Baseline with Guidance, Guidance=true, Treatments=false2026.0417————3——32————
Qwen2.5 7BCategory=Including our Treatments, Guidance=false, Treatments=true2026.0417—104—————————
Llama3.1 8BCategory=Including our Treatments, Guidance=false, Treatments=true2026.0417—14——9———————
WhiteRabbitNeoCategory=Including our Treatments, Guidance=false, Treatments=true2026.041724——————————
Qwen2.5 7BCategory=Secondary Baseline, Guidance=false, Treatments=false2026.048————7———————
Llama3.1 8BCategory=Secondary Baseline, Guidance=false, Treatments=false2026.048————6———————
GPT-4o miniCategory=Secondary Baseline, Guidance=false, Treatments=false2026.048—7——————————
WhiteRabbitNeoCategory=Secondary Baseline, Guidance=false, Treatments=false2026.048—7——————————
Qwen2.5 7BCategory=Secondary Baseline with Guidance, Guidance=true, Treatments=false2026.048————1———————
Llama3 8BCategory=Baseline, Guidance=false, Treatments=false2026.040————————————