Review
Jul 2026
Instruction-Tuned Language Models Cannot Sample from Distributions They Can Describe
For applications that require per-persona outputs, the same model that cannot sample from a distribution can describe it accurately in a single call, and is proposed Prompt-Perturbed Argyle (PPA), which reduces the same error by 21% at no added cost.
Chaemi Jang, Dong-Man Lee, Ji Hee Kim
· arXiv.org · 0 citations