List your company
ResearchFormerly WizWand

Text Classification on HellaSwag

78.9AccuracyGPT-3
42.18851.71961.2570.781Jul 14, 2021Mar 20, 2022Nov 25, 2022Aug 1, 2023Apr 7, 2024Dec 12, 2024Aug 19, 2025
Updated 5mo ago

Evaluation results

MethodLinks
GPT-3# Params=175B, prompting_mode=zero-shot2021.0778.9
I-GLASSBackbone=Mistral 7B, Sparsity=50% FF, Evaluation Protocol=0-shot unnormalized2025.0861.01
GRIFFINBackbone=Mistral 7B, Sparsity=50% FF, Evaluation Protocol=0-shot unnormalized2025.0860.97
GRIFFINBackbone=Gemma 7B, Sparsity=50% FF, Evaluation Protocol=0-shot unnormalized2025.0860.62
I-GLASSBackbone=Gemma 7B, Sparsity=50% FF, Evaluation Protocol=0-shot unnormalized2025.0860.59
I-GLASSBackbone=Llama2 7B, Sparsity=50% FF, Evaluation Protocol=0-shot unnormalized2025.0857.14
GRIFFINBackbone=Llama2 7B, Sparsity=50% FF, Evaluation Protocol=0-shot unnormalized2025.0857.14
I-GLASSBackbone=ReLU-Llama2 7B, Sparsity=50% FF, Evaluation Protocol=0-shot unnormalized2025.0853.95
GRIFFINBackbone=ReLU-Llama2 7B, Sparsity=50% FF, Evaluation Protocol=0-shot unnormalized2025.0853.95
GPT-3 Large# Params=760M, prompting_mode=zero-shot2021.0751
GRIFFINBackbone=OPT 6.7B, Sparsity=50% FF, Evaluation Protocol=0-shot unnormalized2025.0850.49
I-GLASSBackbone=OPT 6.7B, Sparsity=50% FF, Evaluation Protocol=0-shot unnormalized2025.0850.45
HTLM-Manual# Params=400M, prompting_mode=zero-shot2021.0747.9
GPT-3 Med# Params=350M, prompting_mode=zero-shot2021.0743.6