List your company

Exfiltration on ACIARENA Code

51.11UALLM Debate
3.73816.036528.33540.6335Apr 9, 2026

Evaluation results

MethodLinks
LLM DebateBase LLM=GPT-4o-mini2026.0451.1180.22
Self ConsistencyBase LLM=GPT-4o-mini2026.0447.7880
AgentVerseBase LLM=GPT-4o-mini2026.0447.3380.45
AutoGen + ACI-SENTINELMAS Platform=AutoGen, Defense=ACI-SENTINEL2026.0445.110.22
AutoGenBase LLM=GPT-4o-mini2026.0444.8977.55
AgentVerse + ACI-SENTINELMAS Platform=AgentVerse, Defense=ACI-SENTINEL2026.0443.119.11
MetaGPT + SANDWICHMAS Platform=MetaGPT, Defense=SANDWICH2026.0441.7871.56
AgentVerse + SANDWICHMAS Platform=AgentVerse, Defense=SANDWICH2026.0440.8958
AgentVerse + BERT DETECTORMAS Platform=AgentVerse, Defense=BERT DETECTOR2026.0440.6761.11
AgentVerseMAS Platform=AgentVerse, Defense=None2026.044061.56
AutoGen + SANDWICHMAS Platform=AutoGen, Defense=SANDWICH2026.0439.5660
MetaGPTBase LLM=GPT-4o-mini2026.0438.8980.22
AgentVerse + AGRAILMAS Platform=AgentVerse, Defense=AGRAIL2026.0437.7867.33
MetaGPTMAS Platform=MetaGPT, Defense=None2026.0434.6760.44
AgentVerse + G-SAFEGUARDMAS Platform=AgentVerse, Defense=G-SAFEGUARD2026.0433.3356.22
AutoGenMAS Platform=AutoGen, Defense=None2026.0432.2254
MetaGPT + ACI-SENTINELMAS Platform=MetaGPT, Defense=ACI-SENTINEL2026.0428.229.78
MetaGPT + G-SAFEGUARDMAS Platform=MetaGPT, Defense=G-SAFEGUARD2026.042842.89
MetaGPT + DELIMITERMAS Platform=MetaGPT, Defense=DELIMITER2026.0424.8959.78
AgentVerse + DELIMITERMAS Platform=AgentVerse, Defense=DELIMITER2026.0424.2259.56
AutoGen + BERT DETECTORMAS Platform=AutoGen, Defense=BERT DETECTOR2026.0422.6736.67
AutoGen + G-SAFEGUARDMAS Platform=AutoGen, Defense=G-SAFEGUARD2026.0421.3334
AutoGen + DELIMITERMAS Platform=AutoGen, Defense=DELIMITER2026.0420.2244.22
MetaGPT + BERT DETECTORMAS Platform=MetaGPT, Defense=BERT DETECTOR2026.0420.2260.67
CAMELBase LLM=GPT-4o-mini2026.0416.2226
AutoGen + AGRAILMAS Platform=AutoGen, Defense=AGRAIL2026.041429.33
MetaGPT + AGRAILMAS Platform=MetaGPT, Defense=AGRAIL2026.045.5619.78