List your company

Tool-use agent performance on MetaTool

85.9ASR (Success Rate)MemMorph
5.40426.30247.268.098May 24, 2026

Evaluation results

MethodLinks
MemMorphAgent Backbone LLM=Claude 3.5 Opus, Poisoning Type=Direct2026.0585.9
MemMorphAgent Backbone LLM=GPT-4o, Poisoning Type=Direct2026.0583.8
MemMorphAgent Backbone LLM=Llama3 70B, Poisoning Type=Direct2026.0581.2
MemMorphAgent Backbone LLM=Qwen2.5 7B, Poisoning Type=Direct2026.0579.5
MemMorphAgent Backbone LLM=Llama3 8B, Poisoning Type=Direct2026.0578.6
MemMorphAgent Backbone LLM=Claude 3.5 Sonnet, Poisoning Type=Direct2026.0578.6
GCGAgent Backbone LLM=GPT-4o, Poisoning Type=Direct2026.0577.8
MemMorphAgent Backbone LLM=Claude 3.5 Haiku, Poisoning Type=Direct2026.0577.6
GCGAgent Backbone LLM=Claude 3.5 Opus, Poisoning Type=Direct2026.0576.9
PoisonedRAGAgent Backbone LLM=GPT-4o, Poisoning Type=Direct2026.0575.2
MemMorphAgent Backbone LLM=GPT-4o mini, Poisoning Type=Direct2026.0572.7
GCGAgent Backbone LLM=Claude 3.5 Sonnet, Poisoning Type=Direct2026.0572.6
PoisonedRAGAgent Backbone LLM=Claude 3.5 Opus, Poisoning Type=Direct2026.0572.6
MemMorphAgent Backbone LLM=GPT-3.5 20B, Poisoning Type=Direct2026.0572.6
PoisonedRAGAgent Backbone LLM=Claude 3.5 Sonnet, Poisoning Type=Direct2026.0570.9
GCGAgent Backbone LLM=GPT-3.5 20B, Poisoning Type=Direct2026.0570.1
MemMorphAgent Backbone LLM=Qwen2.5 32B, Poisoning Type=Direct2026.0570.1
PoisonedRAGAgent Backbone LLM=GPT-3.5 20B, Poisoning Type=Direct2026.0568.4
GCGAgent Backbone LLM=Llama3 70B, Poisoning Type=Direct2026.0567.8
PoisonedRAGAgent Backbone LLM=GPT-4o mini, Poisoning Type=Direct2026.0563.2
PoisonedRAGAgent Backbone LLM=Llama3 70B, Poisoning Type=Direct2026.0562.4
GCGAgent Backbone LLM=GPT-4o mini, Poisoning Type=Direct2026.0558.1
ToolHijackerAgent Backbone LLM=Claude 3.5 Sonnet, Poisoning Type=Direct2026.0557.3
ToolHijackerAgent Backbone LLM=GPT-3.5 20B, Poisoning Type=Direct2026.0557.2
ToolCommanderAgent Backbone LLM=GPT-4o mini, Poisoning Type=Direct2026.0556.3
PoisonedRAGAgent Backbone LLM=Qwen2.5 32B, Poisoning Type=Direct2026.0555.6
PoisonedRAGAgent Backbone LLM=Claude 3.5 Haiku, Poisoning Type=Direct2026.0555.6
ToolCommanderAgent Backbone LLM=GPT-3.5 20B, Poisoning Type=Direct2026.0555.6
GCGAgent Backbone LLM=Qwen2.5 32B, Poisoning Type=Direct2026.0554.7
ToolHijackerAgent Backbone LLM=GPT-4o, Poisoning Type=Direct2026.0554.4
GCGAgent Backbone LLM=Llama3 8B, Poisoning Type=Direct2026.0553
GCGAgent Backbone LLM=Claude 3.5 Haiku, Poisoning Type=Direct2026.0553
PoisonedRAGAgent Backbone LLM=Llama3 8B, Poisoning Type=Direct2026.0550.4
GCGAgent Backbone LLM=Qwen2.5 7B, Poisoning Type=Direct2026.0549.6
PoisonedRAGAgent Backbone LLM=Qwen2.5 7B, Poisoning Type=Direct2026.0545.3
ToolHijackerAgent Backbone LLM=Claude 3.5 Haiku, Poisoning Type=Direct2026.0544.4
ToolCommanderAgent Backbone LLM=GPT-4o, Poisoning Type=Direct2026.0544.3
ToolHijackerAgent Backbone LLM=Qwen2.5 7B, Poisoning Type=Direct2026.0544.1
ToolHijackerAgent Backbone LLM=GPT-4o mini, Poisoning Type=Direct2026.0541.9
ToolHijackerAgent Backbone LLM=Claude 3.5 Opus, Poisoning Type=Direct2026.0535.8
ToolHijackerAgent Backbone LLM=Qwen2.5 32B, Poisoning Type=Direct2026.0532.1
ToolHijackerAgent Backbone LLM=Llama3 70B, Poisoning Type=Direct2026.0530.7
ToolCommanderAgent Backbone LLM=Claude 3.5 Opus, Poisoning Type=Direct2026.0525.8
ToolHijackerAgent Backbone LLM=Llama3 8B, Poisoning Type=Direct2026.0525.6
ToolCommanderAgent Backbone LLM=Qwen2.5 32B, Poisoning Type=Direct2026.0524.8
ToolCommanderAgent Backbone LLM=Llama3 70B, Poisoning Type=Direct2026.0523
ToolCommanderAgent Backbone LLM=Qwen2.5 7B, Poisoning Type=Direct2026.0516.2
ToolCommanderAgent Backbone LLM=Claude 3.5 Sonnet, Poisoning Type=Direct2026.0515.4
ToolCommanderAgent Backbone LLM=Claude 3.5 Haiku, Poisoning Type=Direct2026.0511.9
ToolCommanderAgent Backbone LLM=Llama3 8B, Poisoning Type=Direct2026.058.5