Content
@
https://warpcast.com/~/channel/aichannel
0 reply
0 recast
0 reaction
shoni.eth
@alexpaden
Prompt Obfuscation - [https://arxiv.org/pdf/2409.11026] Fine tune the model on a hidden language. Option 1: The prompt as an embedding with a model trained on the meaning Option 2: The model trained on a secret language
1 reply
0 recast
2 reactions
shoni.eth
@alexpaden
The purpose is to prevent people from finding system prompts that one might consider as proprietary value.
0 reply
0 recast
0 reaction
