List your company
ResearchFormerly WizWand

Sentiment Analysis on SST-2 (top-1 accuracy)

97Top-1 Accuracy (SST-2)ReLoRA
49.36861.73474.186.466Sep 26, 2025Nov 4, 2025Dec 14, 2025Jan 23, 2026Mar 4, 2026Apr 13, 2026May 23, 2026
Updated 5mo ago

Evaluation results

MethodLinks
ReLoRAUpdate Source=OpenOrca, Evaluation Protocol=Task-adaptation Methods2026.0597
ReLoRAUpdate Source=OpenPlatypus, Evaluation Protocol=Task-adaptation Methods2026.0597
ReLoRAUpdate Source=AlpacaGPT4, Evaluation Protocol=Task-adaptation Methods2026.0596.9
PortLLM+FTUpdate Source=OpenOrca, Evaluation Protocol=Task-adaptation Methods2026.0596.6
ORALUpdate Source=OpenOrca, Evaluation Protocol=Task-adaptation Methods2026.0596.4
PortLLM+FTUpdate Source=AlpacaGPT4, Evaluation Protocol=Task-adaptation Methods2026.0596.4
PortLLM+FTUpdate Source=OpenPlatypus, Evaluation Protocol=Task-adaptation Methods2026.0596.4
LoRA-ScratchUpdate Source=OpenOrca, Evaluation Protocol=Task-adaptation Methods2026.0596.3
ORALUpdate Source=OpenPlatypus, Evaluation Protocol=Task-adaptation Methods2026.0596.2
LoRA-ScratchUpdate Source=AlpacaGPT4, Evaluation Protocol=Task-adaptation Methods2026.0596.1
LoRA-ScratchUpdate Source=OpenPlatypus, Evaluation Protocol=Task-adaptation Methods2026.0596.1
ORALUpdate Source=AlpacaGPT4, Evaluation Protocol=Task-adaptation Methods2026.0596
ZO Fine-tunerBackbone=LLaMA-3.1-8B2025.1094
ZO Fine-tunerBackbone=Qwen2.5-14B2025.1094
FOBackbone=Qwen2.5-14B2025.1094
FOBackbone=LLaMA-3.2-1B2025.1093
FOBackbone=LLaMA-3.1-8B2025.1093
MeZOBackbone=LLaMA-3.1-8B2025.1092
ZO Fine-tunerBackbone=LLaMA-3.2-1B2025.1092
ICRBackbone=Qwen2.5-7B2025.0991
MeZOBackbone=LLaMA-3.2-1B2025.1090
MeZOBackbone=Qwen2.5-14B2025.1088
ICVBackbone=Qwen2.5-7B2025.0983
ELICITBackbone=Qwen2.5-7B2025.0978.5
IVBackbone=Qwen2.5-7B2025.0978.4
TVBackbone=Qwen2.5-7B2025.0978.2
FVBackbone=Qwen2.5-7B2025.0976.8
WUDI-Merging + E-PMQMerging Strategy=WUDI-Merging, Quantization Method=E-PMQ, Bit-width=4-bit, Backbone=CLIP-ViT-B/322026.0569.52
WUDI-MergingMerging Strategy=WUDI-Merging, Quantization Method=Full-precision, Bit-width=Full, Backbone=CLIP-ViT-B/322026.0569
WUDI-Merging + GPTQMerging Strategy=WUDI-Merging, Quantization Method=GPTQ, Bit-width=4-bit, Backbone=CLIP-ViT-B/322026.0567.76
WUDI-Merging + RTNMerging Strategy=WUDI-Merging, Quantization Method=RTN, Bit-width=4-bit, Backbone=CLIP-ViT-B/322026.0567.49
WUDI-Merging + AWQMerging Strategy=WUDI-Merging, Quantization Method=AWQ, Bit-width=4-bit, Backbone=CLIP-ViT-B/322026.0567
TIES-Merging + E-PMQMerging Strategy=TIES-Merging, Quantization Method=E-PMQ, Bit-width=4-bit, Backbone=CLIP-ViT-B/322026.0565.79
Simple Averaging + E-PMQMerging Strategy=Simple Averaging, Quantization Method=E-PMQ, Bit-width=4-bit, Backbone=CLIP-ViT-B/322026.0565.57
Task Arithmetic + E-PMQMerging Strategy=Task Arithmetic, Quantization Method=E-PMQ, Bit-width=4-bit, Backbone=CLIP-ViT-B/322026.0564.74
Simple AveragingMerging Strategy=Simple Averaging, Quantization Method=Full-precision, Bit-width=Full, Backbone=CLIP-ViT-B/322026.0561.8
Simple Averaging + GPTQMerging Strategy=Simple Averaging, Quantization Method=GPTQ, Bit-width=4-bit, Backbone=CLIP-ViT-B/322026.0558.59
Pretrained LLM ΘEvaluation Protocol=Zero-shot2026.0556.9
Simple Averaging + RTNMerging Strategy=Simple Averaging, Quantization Method=RTN, Bit-width=4-bit, Backbone=CLIP-ViT-B/322026.0556.89
Simple Averaging + AWQMerging Strategy=Simple Averaging, Quantization Method=AWQ, Bit-width=4-bit, Backbone=CLIP-ViT-B/322026.0556.84
TIES-Merging + GPTQMerging Strategy=TIES-Merging, Quantization Method=GPTQ, Bit-width=4-bit, Backbone=CLIP-ViT-B/322026.0556.18
Evolved LLM Θ′Update Source=OpenOrca, Evaluation Protocol=Zero-shot2026.0556
TIES-MergingMerging Strategy=TIES-Merging, Quantization Method=Full-precision, Bit-width=Full, Backbone=CLIP-ViT-B/322026.0555.9
Task Arithmetic + GPTQMerging Strategy=Task Arithmetic, Quantization Method=GPTQ, Bit-width=4-bit, Backbone=CLIP-ViT-B/322026.0555.02
TIES-Merging + AWQMerging Strategy=TIES-Merging, Quantization Method=AWQ, Bit-width=4-bit, Backbone=CLIP-ViT-B/322026.0553.82
TIES-Merging + RTNMerging Strategy=TIES-Merging, Quantization Method=RTN, Bit-width=4-bit, Backbone=CLIP-ViT-B/322026.0553.43
Task Arithmetic + AWQMerging Strategy=Task Arithmetic, Quantization Method=AWQ, Bit-width=4-bit, Backbone=CLIP-ViT-B/322026.0552.28
Task Arithmetic + RTNMerging Strategy=Task Arithmetic, Quantization Method=RTN, Bit-width=4-bit, Backbone=CLIP-ViT-B/322026.0552.22
Evolved LLM Θ′Update Source=OpenPlatypus, Evaluation Protocol=Zero-shot2026.0552.2
Evolved LLM Θ′Update Source=AlpacaGPT4, Evaluation Protocol=Zero-shot2026.0551.6
Task ArithmeticMerging Strategy=Task Arithmetic, Quantization Method=Full-precision, Bit-width=Full, Backbone=CLIP-ViT-B/322026.0551.2