List your company

Reasoning on AIME 120

77.5AccuracyReSET
38.08448.31758.5568.783Jun 11, 2026

Evaluation results

MethodLinks
ReSETModel Backbone=Qwen3, Model Size=32B2026.0677.5
BF16 BaselineModel Backbone=Qwen3, Model Size=14B2026.0676.1
BF16 BaselineModel Backbone=Qwen3, Model Size=32B2026.0675.8
4/6Model Backbone=Qwen3, Model Size=32B2026.0674.8
RTNModel Backbone=Qwen3, Model Size=32B2026.0674.4
MR-GPTQModel Backbone=Qwen3, Model Size=32B2026.0673.3
BRQModel Backbone=Qwen3, Model Size=32B2026.0673
ReSETModel Backbone=Qwen3, Model Size=14B2026.0672.1
MR-GPTQModel Backbone=Qwen3, Model Size=14B2026.0671
BF16 BaselineModel Backbone=Qwen3, Model Size=8B2026.0670.4
RTNModel Backbone=Qwen3, Model Size=14B2026.0670.4
4/6Model Backbone=Qwen3, Model Size=14B2026.0670.1
BRQModel Backbone=Qwen3, Model Size=14B2026.0666.9
MR-GPTQModel Backbone=Qwen3, Model Size=8B2026.0665.2
BF16 BaselineModel Backbone=All Models, Model Size=Avg2026.0665.1
ReSETModel Backbone=Qwen3, Model Size=8B2026.0664.9
4/6Model Backbone=Qwen3, Model Size=8B2026.0664
RTNModel Backbone=Qwen3, Model Size=8B2026.0662.5
ReSETModel Backbone=All Models, Model Size=Avg2026.0662.5
4/6Model Backbone=All Models, Model Size=Avg2026.0660.6
MR-GPTQModel Backbone=All Models, Model Size=Avg2026.0660
RTNModel Backbone=All Models, Model Size=Avg2026.0659.9
BF16 BaselineModel Backbone=R1-Qwen, Model Size=14B2026.0657.4
BRQModel Backbone=All Models, Model Size=Avg2026.0657
ReSETModel Backbone=R1-Qwen, Model Size=14B2026.0654
BRQModel Backbone=Qwen3, Model Size=8B2026.0653.8
4/6Model Backbone=R1-Qwen, Model Size=14B2026.0653.1
RTNModel Backbone=R1-Qwen, Model Size=14B2026.0652.4
MR-GPTQModel Backbone=R1-Qwen, Model Size=14B2026.0650.6
BRQModel Backbone=R1-Qwen, Model Size=14B2026.0649.8
BF16 BaselineModel Backbone=R1-Qwen, Model Size=7B2026.0645.7
ReSETModel Backbone=R1-Qwen, Model Size=7B2026.0643.8
BRQModel Backbone=R1-Qwen, Model Size=7B2026.0641.4
4/6Model Backbone=R1-Qwen, Model Size=7B2026.0641.1
RTNModel Backbone=R1-Qwen, Model Size=7B2026.0639.6
MR-GPTQModel Backbone=R1-Qwen, Model Size=7B2026.0639.6