Authors
Loading...
This preprint proposes a geometric interpretation of hallucination in large language models, suggesting structural limits on expressiveness.
Franky Schaut (2026) studied this question.