2 ms·
Makes me wonder if "I don't know" could be added to LLM: whenever an activation has no clear winner value (layman here), couldn't this indicate low response qua
by yukIttEft 2y ago
Makes me wonder if "I don't know" could be added to LLM: whenever an activation has no clear winner value (layman here), couldn't this indicate low response quality?
- Regic 2y agoThis exists and does work to some degree, e.g. Detecting hallucinations in large language models using semantic entropy https://www.nature.com/articles/s41586-024-07421-0 https://www.nature.com/articles/s41586-024-07421-0