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

Mathematical Reasoning on AIME24 (test)

74.27Pass@1 Score

w/o Merging

Updated 9d ago

Evaluation Results

Method	Links
w/o Merging 2026.02		74.27	-
SCF-RKL 2026.02		73.96	-
w/o Merging 2026.02		72.5	-
SCE 2026.02		65.62	-
Ties Merging 2026.02		63.23	-
Dare Task Arithmetic 2026.02		62.92	-
Task Arithmetic 2026.02		61.98	-
Dare Ties Merging 2026.02		61.88	-
RAgent-QwQ-32B 2025.01		56.7	-
Search-o1 2025.01		56.7	-
QwQ-32B 2025.01		53.3	-
DeepSeek-R1-Lite 2025.01		52.5	-
RAG-QwQ-32B 2025.01		50	-
o1-preview 2025.01		44.6	-
DeepSeek-R1-Distill-Qwen-7B 2026.04		43.3	7,891.9
Llama3.3-70B 2025.01		36.7	-
Task Arithmetic 2026.04		36.7	6,580.9
Evo-L2S (Accuracy) 2026.04		36.7	6,674.3
Evo-L2S-01 2026.04		36.7	6,674.3
Evo-L2S-02 2026.04		33.3	4,948.4
RAG-Qwen2.5-32B 2025.01		30	-
ACM-TA 2026.04		30	3,898.1
Evo-L2S-03 2026.04		30	3,140.9
Qwen2.5-32B 2025.01		23.3	-
TIES-Merging 2026.04		23.3	4,266.5
Evo-L2S-04 2026.04		23.3	3,368.4
Qwen2.5-Coder-32B 2025.01		20	-
Qwen2.5-72B 2025.01		20	-
RAgent-Qwen2.5-32B 2025.01		20	-
ACM-Average 2026.04		20	2,951.5
Evo-L2S-05 2026.04		20	2,006.6
Average Merging 2026.04		16.7	3,710.7
ACM-TIES 2026.04		16.7	3,118.6
Evo-L2S (Length) 2026.04		16.7	2,038.9
GPT-4o 2025.01		9.3	-
Qwen2.5-Math-7B 2026.04		0	1,450