Learning Math Reasoning from Self-Sampled Correct and Partially-Correct Solutions
Open paper
Benchmark
Model
Rank
Results
arithmetic-reasoning-on-gsm8k
GPT-Neo-2.7B + Self-Sampling
#265
Accuracy: 19.5
Parameters (Billion): 2.7
arithmetic-reasoning-on-gsm8k
GPT-Neo 125M + Self-Sampling
#274
Accuracy: 7.5
Parameters (Billion): 0.125
Rank counts only results with a code link.