LLMs’ “Not X, but Y” Habit Sparks Debate
Observers have noted that LLMs disproportionately favor the “not X, but Y” construction, sparking debate over why the pattern appears so often. Posts say a common explanation has been challenged, while a researcher has even assigned a new PhD student to quantify the phenomenon empirically.
2026-07-13 ~ 2026-07-15 · 3 related posts
- Why Do AI Models Love the "Not X, But Y" Structure? — TuhinChakr · 2026-07-13
- Why Do LLMs Love Negative Parallelism? — Sauers_ · 2026-07-13
- LLMs Love Saying "It's Not X, It's Y" — TuhinChakr · 2026-07-15