Learning to Build the Environment: Self-Evolving Reasoning RL via Verifiable Environment Synthesis
Open paper
Benchmark
Model
Rank
Results
mathematical-reasoning-on-aime-2025
EvoEnv (Qwen3-4B-Thinking-2507)
–
Accuracy: 83.0
mathematical-reasoning-on-aime24
EvoEnv (Qwen3-4B-Thinking-2507)
–
Acc: 86.2
Rank counts only results with a code link.