RESEARCHMonitorWATCHLIST
Query Timing Produces Opposite Positional Biases Between LLMs and Humans
arXiv cs.CL — Computation and Language
Factual evidence
What the source reports
Academic research reveals LLMs and humans exhibit opposite positional biases depending on when evaluations are requested relative to evidence.
Inspect the evidence
- Inclusion basis
- Enterprise AI
- Publisher and source type
- arXiv cs.CL — Computation and Language · RESEARCH
- Published by source
- 18 November 2026
- Collected by OneBench
- 14 Aug 2026, 15:49 UK
Stored source excerpt
arXiv:2608.12387v1 Announce Type: new Abstract: Positional biases such as recency and primacy effects have been documented in large language models (LLMs), yet the underlying mechanism…
Short excerpt from the collected text, not the full source. Use the source link to read it in context.
The factual summary is a OneBench synthesis, not a quotation or independent verification. Collection time is not publication time. Open the source for its full context; related reporting can share the same underlying announcement.