The WMDP Benchmark: Measuring and Reducing Malicious Use With Unlearning
Paper
•
2403.03218
•
Published
•
1
WMDP Benchmark
The WMDP Benchmark: Measuring and Reducing Malicious Use With Unlearning
-
-
cais/wmdp
Viewer • Updated • 3.67k • 7.17k • 20 -
cais/wmdp-bio-forget-corpus
Viewer • Updated • 24.5k • 403 -
cais/wmdp-cyber-forget-corpus
Viewer • Updated • 1k • 130 • 1 -
cais/wmdp-corpora
Viewer • Updated • 66.4k • 505 • 3 -
cais/wmdp-mmlu-auxiliary-corpora
Viewer • Updated • 8.88k • 39 • 2 -
cais/Zephyr_RMU
Text Generation • 7B • Updated • 83 • 3 -
cais/Mixtral-8x7B-Instruct_RMU
Text Generation • 47B • Updated • 5 • 2 -
cais/Yi-34B-Chat_RMU
Text Generation • 34B • Updated • 5