Share your thoughts, 1 month free Claude Pro on usSee more

Accuracy on AIME 24 (Reasoning)

86.3Accuracy on AIME 24

Qwen3-Base SAT

Updated 26d ago

Evaluation Results

Method	Links
Qwen3-Base SAT 2026.04		86.3
TF-TTCL 2026.04		83.33
Qwen3 (thinking) 2026.04		81.4
Qwen3-A3B 2026.04		80.4
AWQ 2025.11		80
TF-GRPO 2026.04		80
OpenAI o3-mini (medium) 2026.04		79.6
QwQ 2026.04		79.5
ParoQuant 2025.11		77.8
QTIP 2025.11		76.7
TF-TTCL 2026.04		76.67
FP16 2025.11		75.6
FP16 2025.11		75.6
ParoQuant 2025.11		75.6
ParoQuant 2025.11		73.3
FP16 2025.11		73.3
QTIP 2025.11		72.2
AWQ 2025.11		72.2
E-QAT 2025.11		71.1
Qwen3-Base Debate + Judge 2026.04		71.1
Llama 3.1-Base SAT 2026.04		70.3
Chain-of-Thought 2026.04		70
TF-GRPO 2026.04		70
DeepSeek-R1 Distill Llama 2026.04		70
Qwen3-Base Role-play 2026.04		69.9
E-QAT 2025.11		68.9
Qwen3-Base Debate 2026.04		68.7
QTIP 2025.11		67.8
Base LLM 2026.04		66.67
DS-Qwen 14B 2025.05		63.33
AWQ 2025.11		62.2
Layer-wise Caprese 2025.05		61.67
CATS 2025.05		58.34
E2E Caprese 2025.05		58.34
E-QAT 2025.11		53.3
DS-Qwen 7B 2025.05		47.5
FP16 2025.11		42.2
Layer-wise Caprese 2025.05		41.67
Qwen3-Base-DAPO 2026.04		41.3
Chain-of-Thought 2026.04		40
E2E Caprese 2025.05		39.17
Qwen3-Base-GRPO 2026.04		39.1
QTIP 2025.11		37.8
ParoQuant 2025.11		36.6
Layer-wise Caprese 2025.05		35
AWQ 2025.11		34.4
CATS 2025.05		34.17
E2E Caprese 2025.05		33.33
GRIFFIN 2025.05		32.5
E-QAT 2025.11		32.2
Qwen3-Base 2026.04		31.7
Qwen3-Base 2026.04		31
DS-Qwen 1.5B 2025.05		30
E2E Caprese 2025.05		30
Base LLM 2026.04		30
Layer-wise Caprese 2025.05		29.17
GRIFFIN 2025.05		21.67
E2E Caprese 2025.05		20.83
Qwen2.5-Base 2026.04		18.9
Layer-wise Caprese 2025.05		15.83
CATS 2025.05		11.67
GPT-4o-mini-2024-07-18 2026.04		8.1
E2E Caprese 2025.05		6.67
GRIFFIN 2025.05		2.5
Layer-wise Caprese 2025.05		2.5