List your company

Math Problem Solving on GSM8k, SAT-Math, and MATH (AGIEval sampled test)

32.22GSM8k AccuracyCRITIQ
23.151225.505627.8630.2144Feb 26, 2025

Evaluation results

MethodLinks
CRITIQBackbone=Llama-3.2-3B, Continual Pretraining=true, Data Selection Strategy=CRITIQ Scorer2025.0232.2239.556.3426.04
OWMBackbone=Llama-3.2-3B, Continual Pretraining=true, Data Selection Strategy=OpenWebMath subset2025.0228.5132.275.822.19
RawBackbone=Llama-3.2-3B, Continual Pretraining=false2025.0227.6355.522.7
QR-EduBackbone=Llama-3.2-3B, Continual Pretraining=true, Data Selection Strategy=QuRators (Educational Value)2025.0223.536.366.222.02