richardyoung/code-safety-benchmark-results download history
richardyoung/code-safety-benchmark-results is a text classification dataset on the Hugging Face Hub. In the last 30 days it was downloaded 55 times (20 in the last 7 days), and 126 times in total. It ranks #244,111 among datasets by monthly downloads.
Code-Safety Benchmark Results (Paper 2) Per-model classifications and summary statistics for the 13-model behavioral benchmark of Paper 2, "Is Code Safety a Separate Direction? A 13-Model Behavioral Benchmark of Malicious-Code Refusal in Coding-Specialized LLMs" (Young & Moody, 2026).
Open richardyoung/code-safety-benchmark-results on Hugging Face
Sister project: Paper Pulse, the upvote history of every Hugging Face Daily Paper.