List your company

Score Inflation Attack on Chatbot Benchmarks AlpacaEval ArenaHard

2.01Qwen ScoreBITE
0.566480.941241.3161.69076May 24, 2026

Evaluation results

MethodLinks
BITECategory=Ours2026.052.011.9091.3471.7311.356
Fake CompletionCategory=Prompt Injection2026.051.2871.2141.081.0891.103
PAIRCategory=Jailbreak2026.051.2841.8561.2331.3370.869
Escape ChrCategory=Prompt Injection2026.051.2740.8580.961.1861.083
AutoDANCategory=Jailbreak2026.050.9411.3651.1661.4891.091
Null ModelCategory=Prompt Injection2026.050.8780.4230.7411.0921.056
NaiveCategory=Prompt Injection2026.050.8331.0160.7630.8620.65
Context IgnoreCategory=Prompt Injection2026.050.650.7130.7640.7250.727
TAPCategory=Jailbreak2026.050.6220.8710.6560.6930.519