ReflexiCoder: Teaching Large Language Models to Self-Reflect on Generated Code and Self-Correct It via Reinforcement Learning

Open paper
Benchmark Model Rank Results
code-generation-on-humanevalReflexiCoder-8B (Multiple)#13Pass@1: 95.73
code-generation-on-mbppReflexiCoder-8B (Single)#27Accuracy: 81.8

Rank counts only results with a code link.