← 研踪 PaperTrailarXiv:2608.12387 · v2
Query Timing Produces Opposite Positional Biases Between LLMs and Humans
首次提交 2026-08-01 11:23
版本更新 2026-11-17 12:46
雷达观测 2026-11-18 19:43
摘要
Positional biases such as recency and primacy effects have been documented in large language models (LLMs), yet the underlying mechanism by which these models make their evaluations remains poorly understood. Both primacy and recency biases have been observed in human judgments in response to evidence, but recent work suggest that when the listener updates their beliefs - during the presentation of evidence or only at the end - influences the presence of such effects. We investigate whether a similar phenomenon holds for LLMs, finding divergence from human behavior. These biases are more pronounced in newer models compared to their predecessors.
作者备注:EMNLP 2026 Findings; Entropic Award (Top 3 Paper), ICBINB @ ICLR 2026
观测到的版本
相邻记录可能跳过中间版本;首次发现 v2 及以上也会标为版本更新。
v2 · 2026-11-17 12:46 · 最新观测版本
Query Timing Produces Opposite Positional Biases Between LLMs and Humans
Positional biases such as recency and primacy effects have been documented in large language models (LLMs), yet the underlying mechanism by which these models make their evaluations remains poorly understood. Both primacy and recency biases have been observed in human judgments in response to evidence, but recent work suggest that when the listener updates their beliefs - during the presentation of evidence or only at the end - influences the presence of such effects. We investigate whether a similar phenomenon holds for LLMs, finding divergence from human behavior. These biases are more pronounced in newer models compared to their predecessors.
查看此版本解读 →查看此版本原文 ↗