List your company

Instruction Following on ComplexInstruct Level 3

88.7ISRDVRws
31.91646.65861.476.142Oct 16, 2024

Evaluation results

MethodLinks
DVRwsBackbone=Llama-3.1-70B-Instruct-AWQ-INT4, warm-start=true2024.1088.7
DVRcsBackbone=Llama-3.1-70B-Instruct-AWQ-INT4, cold-start=true2024.1088.2
CRITICBackbone=Llama-3.1-70B-Instruct-AWQ-INT42024.1087.6
ReActBackbone=Llama-3.1-70B-Instruct-AWQ-INT42024.1083.5
R-SampleBackbone=Llama-3.1-70B-Instruct-AWQ-INT42024.1083.2
DVRBackbone=Llama3.1-8B, Evaluation Protocol=zero-shot2024.1079.2
GPT-4-turboEvaluation Protocol=zero-shot2024.1078.8
DVRBackbone=Llama-3-8B-Instruct, Start Type=Warm-Start2024.1077
DVRBackbone=Llama-3-8B-Instruct, Start Type=Cold-Start2024.1076.9
CRITICBackbone=Llama-3-8B-Instruct2024.1074.4
ReflexionBackbone=Llama-3.1-70B-Instruct-AWQ-INT42024.1072.8
BSMBackbone=Llama-3.1-70B-Instruct-AWQ-INT42024.1072.6
VanillaBackbone=Llama-3.1-70B-Instruct-AWQ-INT42024.1072.4
U-SCBackbone=Llama-3.1-70B-Instruct-AWQ-INT42024.1071.8
ReActBackbone=Llama-3-8B-Instruct2024.1068.8
DVRBackbone=Mistral-7B, Evaluation Protocol=zero-shot2024.1066.6
R-SampleBackbone=Llama-3-8B-Instruct2024.1064.5
Llama3.1-8BEvaluation Protocol=zero-shot2024.1062.5
ReflexionBackbone=Llama-3-8B-Instruct2024.1057.5
U-SCBackbone=Llama-3-8B-Instruct2024.1056.7
VanillaBackbone=Llama-3-8B-Instruct2024.1056.4
BSMBackbone=Llama-3-8B-Instruct2024.1056
Mistral-7BEvaluation Protocol=zero-shot2024.1034.1