List your company

Arithmetic Reasoning on SVAMP (test)

98.16AccuracySGE
35.697651.913868.1384.3462May 28, 2024Sep 25, 2024Jan 23, 2025May 24, 2025Sep 21, 2025Jan 19, 2026May 20, 2026

Evaluation results

MethodLinks
SGEModel=GPT-4, Tool=Code Interpreter2024.0598.16
Decomp PromptingModel=GPT-4, Tool=Code Interpreter2024.0592.16
CoT PromptingModel=GPT-4, Tool=Code Interpreter2024.0591.58
Refine PromptingModel=GPT-4, Tool=Code Interpreter2024.0591
CoTTest Engine=Phi-3.5-Mini-Instruct2026.0590.4
TextRegTest Engine=Qwen2-7B-Instruct2026.0590.1
CoTTest Engine=Qwen2-7B-Instruct2026.0589.8
TextGradTest Engine=Qwen2-7B-Instruct2026.0589.3
REVOLVETest Engine=Qwen2-7B-Instruct2026.0589.2
LLMBOOSTBackbone=Qwen-2.5-7B, Params=2 × 7B2025.1288.8
TextRegTest Engine=Phi-3.5-Mini-Instruct2026.0588.5
LLMBOOSTBackbone=Qwen-2.5-7B, Params=3 × 7B2025.1288.4
IO PromptingModel=GPT-4, Tool=Code Interpreter2024.0588.34
REVOLVETest Engine=Phi-3.5-Mini-Instruct2026.0587
voteBackbone=Qwen-2.5-7B, Params=3 × 7B2025.1286.7
voteBackbone=Qwen-2.5-7B, Params=2 × 7B2025.1286.4
UNITEBackbone=Qwen-2.5-7B, Params=2 × 7B2025.1286.4
UNITEBackbone=Qwen-2.5-7B, Params=3 × 7B2025.1286.3
Qwen-2.5-7BBackbone=Qwen-2.5-7B, Params=1 × 7B2025.1286.1
UNITEBackbone=Qwen-2.5-3B, Params=3 × 3B2025.1286
LLMBOOSTBackbone=Qwen-2.5-3B, Params=3 × 3B2025.1285.6
CoTTest Engine=Llama-3.1-8B-Instruct2026.0585.5
TextRegTest Engine=Llama-3.1-8B-Instruct2026.0585.5
T-copilotBackbone=Qwen-2.5-7B, Params=2 × 7B2025.1285.3
LLMBOOSTBackbone=Qwen-2.5-3B, Params=2 × 3B2025.1285.3
voteBackbone=Qwen-2.5-3B, Params=3 × 3B2025.1285.2
TextGradTest Engine=Llama-3.1-8B-Instruct2026.0584.8
UNITEBackbone=Qwen-2.5-3B, Params=2 × 3B2025.1284.7
REVOLVETest Engine=Llama-3.1-8B-Instruct2026.0584.6
REVOLVETest Engine=Llama-3-8B-Instruct2026.0583.9
TextRegTest Engine=Llama-3-8B-Instruct2026.0583.9
CoTTest Engine=Llama-3-8B-Instruct2026.0583.7
Qwen-2.5-3BBackbone=Qwen-2.5-3B, Params=1 × 3B2025.1283.4
TextGradTest Engine=Llama-3-8B-Instruct2026.0583.2
T-copilotBackbone=Qwen-2.5-3B, Params=2 × 3B2025.1282.7
voteBackbone=Qwen-2.5-3B, Params=2 × 3B2025.1282.6
LLMBOOSTBackbone=Llama-3.1-8B, Params=3 × 8B2025.1281.3
TextGradTest Engine=Phi-3.5-Mini-Instruct2026.0580.6
LLMBOOSTBackbone=Llama-3.1-8B, Params=2 × 8B2025.1280.1
voteBackbone=Llama-3.1-8B, Params=3 × 8B2025.1278.1
voteBackbone=Llama-3.1-8B, Params=2 × 8B2025.1277.5
Llama-3.1-8BBackbone=Llama-3.1-8B, Params=1 × 8B2025.1277.4
UNITEBackbone=Llama-3.1-8B, Params=3 × 8B2025.1277.2
UNITEBackbone=Llama-3.1-8B, Params=2 × 8B2025.1276.5
T-copilotBackbone=Llama-3.1-8B, Params=2 × 8B2025.1275.7
DoRA + MoDEKBackbone=LLaMA3-8B2024.1074.7
LoRAALL + MoDEKBackbone=LLaMA3-8B2024.1074.2
DoRABackbone=LLaMA3-8B2024.1074
LoRAALLBackbone=LLaMA3-8B2024.1073.8
LoRA¬K + MoDEK (+0.04%)Backbone=LLaMA3-8B2024.1073.5
LoRA¬KBackbone=LLaMA3-8B2024.1072.8
LLMBOOSTBackbone=LLama-3.2-3B, Params=2 × 3B2025.1270.1
LLMBOOSTBackbone=LLama-3.2-3B, Params=3 × 3B2025.1269.9
voteBackbone=LLama-3.2-3B, Params=2 × 3B2025.1268.2
REFTBackbone=LLaMA3-8B2024.1067.5
T-copilotBackbone=LLama-3.2-3B, Params=2 × 3B2025.1266.7
voteBackbone=LLama-3.2-3B, Params=3 × 3B2025.1266.4
UNITEBackbone=LLama-3.2-3B, Params=3 × 3B2025.1266
LLama-3.2-3BBackbone=LLama-3.2-3B, Params=1 × 3B2025.1265.8
UNITEBackbone=LLama-3.2-3B, Params=2 × 3B2025.1265.7
RoAd4Model=13B, #Params.=0.07%2024.0859.5
RoAd2Model=13B, #Params.=0.03%2024.0858.3
RoAd1Model=13B, #Params.=0.02%2024.0857.3
AdapterPModel=13B, #Params.=2.89%2024.0855.7
LORAModel=13B, #Params.=0.67%2024.0854.6
LoReFTModel=13B, #Params.=0.03%2024.0854.2
AdapterSModel=7B, #Params.=0.99%2024.0852.3
LORAModel=7B, #Params.=0.83%2024.0852.1
AdapterSModel=13B, #Params.=0.80%2024.0850.8
AdapterPModel=7B, #Params.=3.54%2024.0849.6
RoAd4Model=7B, #Params.=0.08%2024.0849.4
RoAd2Model=7B, #Params.=0.04%2024.0849.3
(IA)3Model=7B, #Params.=0.02%2024.0848.5
DoRA + MoDEKBackbone=LLaMA2-7B2024.1048.4
DoRABackbone=LLaMA2-7B2024.1048.2
LoRAALL + MoDEKBackbone=LLaMA2-7B2024.1048
LoRAALLBackbone=LLaMA2-7B2024.1047.7
LoRA¬K + MoDEK (+0.04%)Backbone=LLaMA2-7B2024.1047.5
LoReFTModel=7B, #Params.=0.03%2024.0846.8
RoAd1Model=7B, #Params.=0.02%2024.0846.7
LoRA¬KBackbone=LLaMA2-7B2024.1046.6
REFTBackbone=LLaMA2-7B2024.1044
PrefixModel=13B, #Params.=0.03%2024.0841.4
PrefixModel=7B, #Params.=0.04%2024.0838.1