List your company

Multi-turn Instruction Following on StructFlowBench

89.46CSR+GraphIF
63.7270.402577.08583.7675Nov 13, 2025

Evaluation results

MethodLinks
+GraphIFBackbone Model=Qwen2.5-7B-Instruct2025.1189.4669.2589.4788.6
+GraphIFBackbone Model=Llama-3.1-8B-Instruct2025.1188.9170.2989.3888.04
+GraphIFBackbone Model=Llama-3.1-Tulu-3.1-8B2025.1188.6869.7188.9887.9
+GraphIFBackbone Model=Llama-3.1-Storm-8B2025.1186.9264.1886.8385.32
LLM-onlyBackbone Model=Llama-3.1-Tulu-3.1-8B2025.1180.9247.8581.9378.78
+GraphIFBackbone Model=Hermes-3-Llama-3.1-8B2025.1178.7446.7479.7376.54
+MemoryBankBackbone Model=Llama-3.1-Tulu-3.1-8B2025.1178.5142.2478.8875.32
+MemoryBankBackbone Model=Qwen2.5-7B-Instruct2025.1174.6329.9974.9270.58
+MemoryBankBackbone Model=Llama-3.1-8B-Instruct2025.1174.3735.8575.4870.83
LLM-onlyBackbone Model=Llama-3.1-Storm-8B2025.1174.0734.1274.9970.75
+MemoryBankBackbone Model=Llama-3.1-Storm-8B2025.1173.9231.9474.3770.02
+MemoChatBackbone Model=Llama-3.1-8B-Instruct2025.1173.5132.0774.2969.86
+MemoChatBackbone Model=Qwen2.5-7B-Instruct2025.1172.8727.0873.2368.72
+MemoChatBackbone Model=Llama-3.1-Storm-8B2025.1171.5530.671.6267.78
LLM-onlyBackbone Model=Llama-3.1-8B-Instruct2025.1171.0329.3272.7567.98
LLM-onlyBackbone Model=Qwen2.5-7B-Instruct2025.1170.723.0271.2267.07
+MemoryBankBackbone Model=Hermes-3-Llama-3.1-8B2025.1169.223.5770.0364.98
+MemoChatBackbone Model=Hermes-3-Llama-3.1-8B2025.1168.7924.6469.8964.76
LLM-onlyBackbone Model=Hermes-3-Llama-3.1-8B2025.1168.3923.5569.7465.11
+MemoChatBackbone Model=Llama-3.1-Tulu-3.1-8B2025.1164.7125.1366.4661.85