Style Outweighs Substance: Failure Modes of LLM Judges in Alignment Benchmarking
In this reference
A study of how presentation style can bias language-model judges.
Cite the original work
Benjamin Feuer & Micah Goldblum & Teresa Datta & Sanjana Nambiar & Raz Besaleli & Samuel Dooley & Max Cembalest & John P. Dickerson (23 September 2024). Style Outweighs Substance: Failure Modes of LLM Judges in Alignment Benchmarking [Research paper]. arXiv. https://arxiv.org/abs/2409.15268v3
View BibTeX
@misc{style_outweighs_substance_failure_modes_of_llm_judges_in_alignment_benchmarking,
author = {Benjamin Feuer and Micah Goldblum and Teresa Datta and Sanjana Nambiar and Raz Besaleli and Samuel Dooley and Max Cembalest and John P. Dickerson},
title = {{Style Outweighs Substance: Failure Modes of LLM Judges in Alignment Benchmarking}},
howpublished = {arXiv},
url = {https://arxiv.org/abs/2409.15268v3},
year = {2024}
}
