Still No Lie Detector for Language Models: Probing Empirical and Conceptual Roadblocks — mentioned by astrange

Mon, 24 Jul 2023 09:29:37 GMT

> I think the author of that title could well do with a refresher course in epistemology and physics, as it is just not possible to do what they suggest. If you define "facts" as "things actually stored in the LLM's weights", then research shows it is possible to determine if an output is a "fact" or not. Although looking on arxiv I found a paper saying it doesn't work (https://arxiv.org/pdf/2307.00175.pdf) so maybe not.

Paper and discussion history. A mention is not an endorsement.

Still No Lie Detector for Language Models: Probing Empirical and Conceptual Roadblocks · HackerPapers discussions

Still No Lie Detector for Language Models: Probing Empirical and Conceptual Roadblocks — mentioned by astrange