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

Mathematical Reasoning on TabMWP (accuracy)

97.61Accuracy

Zero-shot-EI

Updated 17d ago

Evaluation Results

Method	Links
Zero-shot-EI 2026.02		97.61
ExGRPO 2026.02		97.43
NPG-Muse-8B 2025.08		97.2
ExGRPO 2026.02		97.1
Qwen2.5-14B-Instruct-1M 2025.08		96.9
RevThink 2026.02		96.73
On-Policy Distillation 2026.02		96.15
Answer Aug 2026.02		96.12
GPT-4 2024.05		95.9
GPT-4 (Code Interpreter) 2024.02		95.9
Question Aug 2026.02		95.67
Divide-or-Conquer 2026.02		95.62
Rephrase Question 2026.02		95.12
Zero-shot 2026.02		95.1
Distill Step-by-Step 2026.02		94.76
SKD 2026.02		94.17
Ori-SFT 2025.06		93.9
AbstRaL 2025.06		93.9
Zero-shot 2026.02		93.88
Qwen3-14B-Base 2025.08		93.2
CoA 2025.06		92.9
ExGRPO 2026.02		91.23
GPT-4 2024.05		90.8
ExGRPO 2026.02		90.55
NPG-Muse-7B 2025.08		89.9
CoT-RL 2025.06		89.2
On-Policy Distillation 2026.02		86.21
RevThink 2026.02		84.71
JiuZhang3.0 2024.05		84.7
Qwen3-8B-Base 2025.08		84
ChatGPT 2024.05		82
Qwen2.5-7B (Base) + RL w/ SIE-100% 2025.09		81.8
Qwen-1.5 2024.05		80.5
ChatGPT 2024.05		79.9
JiuZhang3.0-8B 2024.05		79.9
JiuZhang3.0 2024.05		79.2
Qwen2.5-7B-Instruct-1M 2025.08		78.4
Llama3.1-8B-Instruct + RL w/ SIE-100% 2025.09		77
JiuZhang3.0-7B 2024.05		75.6
Divide-or-Conquer 2026.02		75.18
MAmmoTH2 2024.05		75.1
JiuZhang3.0 2024.05		74.8
COMT 2026.01		74.3
OpenMath-CodeLlama 2024.02		74.2
COMT+CCRL 2026.01		74.2
ToRA 2024.02		74
FEW-SHOT 2026.01		73.9
QWEN3-8B 2026.01		73.9
COMT 2026.01		73.6
COMT+CCRL 2026.01		72.8
CoT-SFT 2026.01		71.5
OpenMath-Llama2 2024.02		70.8
DeepSeekMath 2024.05		70.5
ToRA 2024.02		70.5
FEW-SHOT 2026.01		70.4
QWEN2.5-7B 2026.01		70.4
Base LLM 2026.01		70.4
Base LLM 2026.01		70.4
FEW-SHOT 2026.01		70.3
QWEN3-4B 2026.01		70.3
Base LLM 2026.01		70.3
Base LLM 2026.01		70.3
FSLR 2026.01		70.2
Rho-1-Math-7B-Code 2024.05		70.1
OpenMath-Mistral-7B 2024.05		70
OpenMath-Mistral-7B 2024.02		70
CoT-SFT 2026.01		70
ZERO-SHOT 2026.01		69.9
DeepSeek-Math 2026.01		69.9
DeepSeek-Math 2026.01		69.9
DeepSeek-Math 2026.01		69.9
DeepSeekMath 2024.05		69.8
COT-SFT 2026.01		69.7
COMT+CCRL 2026.01		69.6
FSLR 2026.01		69.6
COMT+CCRL 2026.01		69.5
Llama3.1-8B-Instruct + CoT 2025.09		69.5
FSLR 2026.01		69.1
COMT 2026.01		68.6
COMT 2026.01		68.6
COT-SFT+RL 2026.01		68.5
COMT+CCRL 2026.01		68.3
LLAMA-3 2024.05		67.5
Mixtral 2024.05		67.3
FSLR 2026.01		67.2
FSLR 2026.01		67.2
FSLR 2026.01		67
FSLR 2026.01		66.7
CoT-SFT 2026.01		66.6
COMT+CCRL 2026.01		66.4
CoT-SFT 2026.01		66.4
ZERO-SHOT 2026.01		66.3
QWEN3-4B 2026.01		66.3
Base LLM 2026.01		66.3
Base LLM 2026.01		66.3
COT-SFT 2026.01		66.1
OpenMath-CodeLlama 2024.02		66
Answer Aug 2026.02		65.93
COMT 2026.01		65.7
COT-SFT 2026.01		65.7

Showing 100 of 188 rows