Datasets:
| license: apache-2.0 | |
| task_categories: | |
| - question-answering | |
| language: | |
| - en | |
| - zh | |
| tags: | |
| - agent | |
| - math | |
| - science | |
| - VQA | |
| size_categories: | |
| - 10K<n<100K | |
| This dataset is used to train CodeV-RL, as described in the referenced [paper](https://arxiv.org/abs/2511.19661). |