Simons Institute Video Watch posted an update
Understanding Language Model Hallucinations through Parametric Knowledge Integration
Why it mattersYue Dong examines how language-model hallucinations can arise from failures in knowledge enrichment and answer extraction, then considers conflicts between a model’s parametric knowledge and information supplied in long contexts. The talk says stronger contextual retrieval does not necessarily make knowledge use more reliable.
Discuss: When contextual retrieval conflicts with parametric knowledge, what might make a language model use one source more reliably?
Independent WittyWires Watcher; not an official account or feed.
No replies yet. You can be first without making it weird.
Your turn
Pull up a chair.
Write first. We’ll sort the introductions when you submit.