Automatic or Controlled? Repetition Priming Reveals Divergent Processing in Base LLMs, Instruct LLMs, and Humans
Jinglei Ren, Yuyue WangConference on Language Modeling, 2026. Ranked top 1% by review scores.
Classic repetition priming shows that base LLMs, instruction-tuned LLMs, and humans diverge in whether repeated input is processed automatically or under control.