List your company
ResearchFormerly WizWand

General Reasoning on Average of Reasoning Tasks

63.31Average AccuracyPASER
24.382834.488944.59554.7011Feb 18, 2025May 4, 2025Jul 18, 2025Oct 2, 2025Dec 16, 2025Mar 1, 2026May 16, 2026
Updated 5mo ago

Evaluation results

MethodLinks
PASERQuantization=GPTQ 4 bits, Model=LLaMA2-7B2025.0263.31
w/o trainingQuantization=w/o Quant, Model=LLaMA2-7B2025.0262.91
PASERQuantization=RTN 4 bits, Model=LLaMA2-7B2025.0262.58
NuggetsQuantization=GPTQ 4 bits, Model=LLaMA2-7B2025.0261.96
IFDQuantization=GPTQ 4 bits, Model=LLaMA2-7B2025.0261.42
NuggetsQuantization=RTN 4 bits, Model=LLaMA2-7B2025.0261.23
Instruction MiningQuantization=GPTQ 4 bits, Model=LLaMA2-7B2025.0260.68
IFDQuantization=RTN 4 bits, Model=LLaMA2-7B2025.0260.49
Full DataQuantization=GPTQ 4 bits, Model=LLaMA2-7B2025.0260.4
w/o trainingQuantization=GPTQ 4 bits, Model=LLaMA2-7B2025.0260.21
Full DataQuantization=RTN 4 bits, Model=LLaMA2-7B2025.0259.74
RandomQuantization=GPTQ 4 bits, Model=LLaMA2-7B2025.0259.34
Instruction MiningQuantization=RTN 4 bits, Model=LLaMA2-7B2025.0259.23
w/o trainingQuantization=RTN 4 bits, Model=LLaMA2-7B2025.0259.05
RandomQuantization=RTN 4 bits, Model=LLaMA2-7B2025.0257.48
+ SPICEBackbone=Qwen3-8B-Base, #Data=20K2026.0538.53
D2EvoBackbone=Qwen3-8B-Base, #Data=1.7K, Iteration=32026.0536.13
+ Full DataBackbone=Qwen3-8B-Base, #Data=19K2026.0535.61
+ AZRBackbone=Qwen3-8B-Base2026.0535.6
+ R-ZeroBackbone=Qwen3-8B-Base2026.0535.28
Base ModelBackbone=Qwen3-8B-Base2026.0533.54
+ SPICEBackbone=Qwen3-4B-Base, #Data=20K2026.0533.53
D2EvoBackbone=Qwen3-4B-Base, #Data=1.4K, Iteration=32026.0532.16
+ Full DataBackbone=Qwen3-4B-Base, #Data=19K2026.0531.89
+ R-ZeroBackbone=Qwen3-4B-Base2026.0530.24
+ AZRBackbone=Qwen3-4B-Base2026.0529.33
D2EvoBackbone=Llama-3.1-8B, #Data=1.2K, Iteration=32026.0529.07
+ Full DataBackbone=Llama-3.1-8B, #Data=19K2026.0528.39
Base ModelBackbone=Qwen3-4B-Base2026.0527.96
Base ModelBackbone=Llama-3.1-8B2026.0525.88