XTC Sampling Makes Creative Outputs More Diverse Without Hurting Fluency
ziv_ravid · x · 2026-09-26
Experiments across model families show XTC makes creative outputs more diverse and less repetitive, with human raters preferring it for creativity at no cost to fluency.
At equal diversity it hurts instruction following far less than temperature, and it stacks with temperature and min-p.
More from Research
- NYU defines an interestingness metric for theorems and trains a 27B model to discover new math — nyuniversity · 2026-09-26
- Why Absolute Zero Reasoner stands out: a self-play model needing zero external data — cephaloform · 2026-09-26
- Absolute Zero Reasoner revisited: self-play reasoning with zero external data and its 'uh-oh' moment — cephaloform · 2026-09-26
- Should Double-Blind Peer Review Be Replaced by Fully Open Review in the AI Era? — Temporary_Switch_339 · 2026-09-26
- Jev-Mem: System-One/Two-Inspired Memory Architecture Builds Agent Memory 6.6x Faster — omarsar0 · 2026-09-26
- Stanford paper on boosting LLM creativity via cross-domain analogy accepted at NeurIPS 2026 — james_y_zou · 2026-09-26