2 ms·
The idea that OpenAI was intentionally watermarking its output to avoid training data back-contamination should be thoroughly discredited now.
by kmeisthax 3y ago
The idea that OpenAI was intentionally watermarking its output to avoid training data back-contamination should be thoroughly discredited now.