Watermarked LLM outputs look identical to humans
A quiz shows that even technical readers cannot spot which GPT-4 response has a cryptographic watermark baked in. The detection gap is real.
Read the note 1 min read
From the notebook
watermarking.
1 note on this topic.