List your company

Mathematical Reasoning on GSM8K, Minerva, MATH, and MathQA

59.24GSM8K ScoreDataAgentRL
-0.632814.911130.45545.9989Jul 21, 2025

Evaluation results

MethodLinks
DataAgentRLTarget Model=LLaMA-DCLM, GPU Hrs=1891.84, Data Reweighting Space=2-dimensional2025.0759.2414.822.7535.347.03547.66
Naive TrainingModel Backbone=LLaMA-DCLM, Data Reweighting Space=52-dimensional (Nvidia domain classifier)2025.0759.2116.1622.8532.9632.8—
Naive TrainingTarget Model=LLaMA-DCLM2025.0759.2116.1622.8532.9632.8—
Naive TrainingTarget Model=LLaMA-DCLM, GPU Hrs=874.88, Data Reweighting Space=2-dimensional2025.0759.2116.1622.8532.9638.51—
Naive TrainingModel Backbone=LLaMA-Nemo, Data Reweighting Space=52-dimensional (Nvidia domain classifier)2025.0759.0512.324.5232.5332.1—
Naive TrainingTarget Model=LLaMA-Nemo2025.0759.0512.324.5232.5332.1—
Naive TrainingTarget Model=LLaMA-Nemo, GPU Hrs=870.4, Data Reweighting Space=2-dimensional2025.0759.0512.324.5232.5337.86—
DataAgentRLModel Backbone=LLaMA-DCLM, Data Reweighting Space=52-dimensional (Nvidia domain classifier)2025.0759.0416.4822.931.7232.54—
DataAgentRLTarget Model=LLaMA-DCLM2025.0759.0416.4822.931.7232.54—
DataAgentRLModel Backbone=LLaMA-FWE, Data Reweighting Space=52-dimensional (Nvidia domain classifier)2025.0758.9614.6823.6233.3232.65—
DataAgentRLTarget Model=LLaMA-FWE2025.0758.9614.6823.6233.3232.65—
Naive TrainingModel Backbone=LLaMA-FWE, Data Reweighting Space=52-dimensional (Nvidia domain classifier)2025.0758.9112.5824.733.9432.53—
Naive TrainingTarget Model=LLaMA-FWE2025.0758.9112.5824.733.9432.53—
Naive TrainingTarget Model=LLaMA-FWE, GPU Hrs=852.1, Data Reweighting Space=2-dimensional2025.0758.9112.5824.733.9438.51—
DataAgentRLTarget Model=LLaMA-FWE, GPU Hrs=1838.2, Data Reweighting Space=2-dimensional2025.0758.0713.4623.9633.2845.48540.83
DataAgentSFTTarget Model=LLaMA-DCLM, GPU Hrs=1880.32, Data Reweighting Space=2-dimensional2025.0757.8412.721.332.7344.95563.27
DataAgentRLModel Backbone=LLaMA-Nemo, Data Reweighting Space=52-dimensional (Nvidia domain classifier)2025.0756.7811.8223.6132.0331.06—
DataAgentRLTarget Model=LLaMA-Nemo2025.0756.7811.8223.6132.0331.06—
DataAgentSFTModel Backbone=LLaMA-DCLM, Data Reweighting Space=52-dimensional (Nvidia domain classifier)2025.0756.7711.1218.7632.329.74—
DataAgentSFTTarget Model=LLaMA-DCLM2025.0756.7711.1218.7632.329.74—
DataAgentSFTTarget Model=LLaMA-FWE, GPU Hrs=1887.82, Data Reweighting Space=2-dimensional2025.0756.2612.1921.9231.3245.23552.61
DBLTarget Model=LLaMA-DCLM, GPU Hrs=1906.8, Data Reweighting Space=2-dimensional2025.0756.2212.218.2430.1443.5575.91
RegMixTarget Model=LLaMA-FWE, GPU Hrs=1948.26, Data Reweighting Space=2-dimensional2025.0755.912.220.3530.143.83557.18
RegMixTarget Model=LLaMA-DCLM, GPU Hrs=1987.84, Data Reweighting Space=2-dimensional2025.0755.8711.517.732.1644.01559.83
RegMixModel Backbone=LLaMA-DCLM, Data Reweighting Space=52-dimensional (Nvidia domain classifier)2025.0755.7810.3615.7528.6727.64—
RegMixTarget Model=LLaMA-DCLM2025.0755.7810.3615.7528.6727.64—
DataAgentSFTModel Backbone=LLaMA-FWE, Data Reweighting Space=52-dimensional (Nvidia domain classifier)2025.0754.513.2722.1731.130.26—
DataAgentSFTTarget Model=LLaMA-FWE2025.0754.513.2722.1731.130.26—
DataAgentRLTarget Model=LLaMA-Nemo, GPU Hrs=1812.31, Data Reweighting Space=2-dimensional2025.0754.2810.8322.9431.8545.8520.71
DBLTarget Model=LLaMA-FWE, GPU Hrs=1883.58, Data Reweighting Space=2-dimensional2025.0754.212.0520.7231.5243.92566.47
RegMixModel Backbone=LLaMA-FWE, Data Reweighting Space=52-dimensional (Nvidia domain classifier)2025.0753.6810.8420.353028.72—
RegMixTarget Model=LLaMA-FWE2025.0753.6810.8420.353028.72—
DataAgentSFTModel Backbone=LLaMA-Nemo, Data Reweighting Space=52-dimensional (Nvidia domain classifier)2025.0752.510.720.7732.7129.17—
DataAgentSFTTarget Model=LLaMA-Nemo2025.0752.510.720.7732.7129.17—
RegMixTarget Model=LLaMA-Nemo, GPU Hrs=1925.23, Data Reweighting Space=2-dimensional2025.0749.3910.719.233044.13546.24
Naive TrainingModel Backbone=Pythia-1.4B, Data Reweighting Space=52-dimensional (Nvidia domain classifier)2025.0748.9810.1614.6427.5225.33—
RegMixModel Backbone=LLaMA-Nemo, Data Reweighting Space=52-dimensional (Nvidia domain classifier)2025.0748.1710.9119.0330.0427.04—
RegMixTarget Model=LLaMA-Nemo2025.0748.1710.9119.0330.0427.04—
DataAgentSFTTarget Model=LLaMA-Nemo, GPU Hrs=1887.25, Data Reweighting Space=2-dimensional2025.0747.819.0717.828.8644.12538.44
DBLTarget Model=LLaMA-Nemo, GPU Hrs=1920.21, Data Reweighting Space=2-dimensional2025.0745.9110.7420.0128.7442.63522.88
DataAgentRLModel Backbone=Pythia-1.4B, Data Reweighting Space=52-dimensional (Nvidia domain classifier)2025.0745.448.110.0125.3822.23—
DataAgentSFTModel Backbone=Pythia-1.4B, Data Reweighting Space=52-dimensional (Nvidia domain classifier)2025.0738.847.197.320.6818.5—
RegMixModel Backbone=Pythia-1.4B, Data Reweighting Space=52-dimensional (Nvidia domain classifier)2025.0738.287.247.5121.118.53—
Base ModelModel Backbone=LLaMA-DCLM, Data Reweighting Space=52-dimensional (Nvidia domain classifier)2025.072.554.14.2224.528.85—
Base ModelTarget Model=LLaMA-DCLM2025.072.554.14.2224.528.85—
Base ModelTarget Model=LLaMA-DCLM, Data Reweighting Space=2-dimensional2025.072.554.14.2224.5238.15780.62
Base ModelModel Backbone=LLaMA-FWE, Data Reweighting Space=52-dimensional (Nvidia domain classifier)2025.072.52.524.0623.728.2—
Base ModelModel Backbone=LLaMA-Nemo, Data Reweighting Space=52-dimensional (Nvidia domain classifier)2025.072.54.855.323.629.07—
Base ModelTarget Model=LLaMA-FWE2025.072.52.524.0623.728.2—
Base ModelTarget Model=LLaMA-Nemo2025.072.54.855.323.629.07—
Base ModelTarget Model=LLaMA-FWE, Data Reweighting Space=2-dimensional2025.072.52.524.0623.7237.65781.34
Base ModelTarget Model=LLaMA-Nemo, Data Reweighting Space=2-dimensional2025.072.54.855.323.6238.22782.95
Base ModelModel Backbone=Pythia-1.4B, Data Reweighting Space=52-dimensional (Nvidia domain classifier)2025.071.674.392.114.165.58—