TIGER-Lab/EvolveCoder-22K

EvolveCoder-22K Overview EvolveCoder-22K is a large-scale coding dataset for reinforcement learning, built using the EvolveCoder framework. It contains 21,600 competitive programming problems paired with adversarially evolved test cases that are designed to be harder and more discriminative than standard test suites. Each problem comes with a set of test cases generated through multiple rounds of adversarial verification — candidate solutions are run against… See the full description on the dataset page: https://huggingface.co/datasets/TIGER-Lab/EvolveCoder-22K.

<iframe src="https://silytics.si/embed/item/4694262d-d8ac-4c4f-80cf-533153d89d68" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 17 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.