Snooow1029/advbench-audio-inhouse download history

Snooow1029/advbench-audio-inhouse is an audio classification dataset on the Hugging Face Hub. In the last 30 days it was downloaded 149 times (4 in the last 7 days), and 214 times in total. It ranks #85,279 among datasets by monthly downloads.

AdvBench-Audio harmful (in-house TTS) 520 harmful AdvBench behaviors, single-speaker in-house TTS (neutral voice). Used as the SPIRIT noise-sensitive-neuron calibration source (clean vs random-noise activation diff). DISTINCT TTS from WeifeiJin/AdvBench-Audio (that repo ships advbench

Open Snooow1029/advbench-audio-inhouse on Hugging Face

Sister project: Paper Pulse, the upvote history of every Hugging Face Daily Paper.