The study examines how 105 language models choose a single word from a large set of valid options, finding that 46% of the time they all pick the same word, "serendipity." Using 96 prompts that ask for one-word answers across various categories, the researchers measure convergence and each model’s share of the consensus. They observe that post‑training and persona‑tuned models diverge more, while heavily post‑trained assistants are more conformist, and that the field’s answer choices are more concentrated than human norms in most categories.
By Tapan Parikh
The paper introduces the pooled selection-rate standard deviation (pooled SD) as a way to quantify how unevenly multiple-choice option letters are chosen across different orderings. It shows that pooled SD decomposes into two label‑free factors: the flip rate, which measures how often a model’s chosen option changes with ordering, and the positional skew, which captures the variability in option‑letter selection rates. Experiments on 41 checkpoints and 12 benchmarks reveal that prior correction mainly reduces positional skew, while aggregating each ordering with its reversal mainly reduces flip rate; applying both mitigations together further narrows the accuracy range across answer letters.
By Hyunjong Ok, Minjae Park, Jaeho Lee
arXiv:2602.13576v2 Announce Type: replace-cross
Abstract: Evaluation and alignment pipelines for large language models increasingly rely on LLM-based judges, whose behavior is guided by natural-langu...
By Ruomeng Ding, Yifei Pang, He Sun, Yizhong Wang, Zhiwei Steven Wu, Zhun Deng
arXiv:2602.02219v3 Announce Type: replace
Abstract: Large language models are widely employed as evaluators, a paradigm commonly referred to as LLM-as-a-judge. Prior research has predominantly examin...
By Yuzheng Xu, Tosho Hirasawa, Tadashi Kozuno, Yoshitaka Ushiku
arXiv:2610.35826v1 Announce Type: new
Abstract: Selection bias in large language models (LLMs) can persist despite instructions to ignore presentation order or prompts that elicit reasoning. Recent e...
By Mingyang Song, Mao Zheng, Zhiyuan Ma, Xinpeng Zhou
The paper investigates how large language models (LLMs) from OpenAI and Anthropic exhibit probability bias when deciding whether to trust information from reporters of different racial descriptions. Using a two‑urn Bayesian game and 6,600 sessions, the authors find that models often refuse to assess a reporter’s credibility—especially for Black‑described reporters—yet when they do assess, Black reporters receive slightly higher trust. The bias is subtle, growing with reasoning effort, and is not detected by standard audits that compare average trust or only assessed cases.
By Christian Arnold, Ernesto Calvo, Guadalupe Gonzalez, Sebasti\'an Vallejo Vera, Tiago Ventura
The paper introduces a reference‑based bias detection method that audits hidden‑state representations of language models by encoding sentences as similarities to a fixed set of anchor sentences. This relative representation allows comparison across model variants, such as before and after fine‑tuning, and yields a metric called Representational Bias Shift (ΔB). ΔB correlates strongly with output‑level bias changes, can detect bias‑increasing checkpoints with high ROC AUC, and is computationally efficient, requiring only a few minutes and far less compute than traditional benchmarks.
By Marek Jeli\'nski, Jan Dubi\'nski, Maciej Chrabaszcz, Sebastian Cygert
The paper introduces an audit layer for vision‑language model benchmarks that reuses per‑question artifacts from evaluation runs to assess the benchmark itself. Six detectors analyze saved outputs, and an additional text‑only rerun removes visual inputs to produce per‑question reports and aggregate statistics, including permutation‑null excess and interval estimates. Applying this method to the MMStar benchmark reveals that 20.2% of its questions show no accuracy gain from images, with a small but significant excess in mathematics and science categories, and identifies questions where all models chose the same non‑gold option.
By Denis Degtiarev, Anna Gromova, Sergey Katkov, Artem Mikhailin, Safina Mukhamedshina, Albert Nasybullin, Karina Shaikhutdinova, Artem Tuzov
arXiv:2606. 03650v1 Announce Type: cross Abstract: Choosing or ranking language models for a specific application is hardest when no task-specific labeled data exists, and standard public benchmarks cannot be trusted, their items having likely leaked into pretraining, so scores reflect memorization rather than fitness.
By Alexander Apartsin, Yehudit Aperstein
The paper demonstrates that typed decision models can produce type‑correct outputs that nevertheless do not adhere to the explicit definitions of their options. By swapping only the names of options while keeping the definitions, question, and state unchanged, the authors show that decision flips can increase dramatically—up to 70.4 percentage points—when using yes/no names versus neutral controls. This effect is consistent across multiple models and decision rules, and even random strings as names yield flip rates similar to neutral controls, indicating that option‑name polarity contributes to decision instability beyond mere reassignment.
By Yu Sun, Junhao Xu, Jiajia Shi, Zijin Yang
arXiv:2604. 22891v4 Announce Type: replace-cross Abstract: LLM-as-a-Judge has become a dominant approach in automated evaluation systems, playing critical roles in model alignment, leaderboard construction, quality control, and so on.
By Jinming Yang, Zheng Hu, Chuxian Qiu, Zhenyu Deng, Xinshan Jiao, Tao Zhou
arXiv:2610.17641v1 Announce Type: new
Abstract: Bias scores in language-model evaluations often conflate what a model prefers with whether it answers at all. We identify this confound in \textsc{BBQ}...
By Weitong Li, Donglin Song, Junkai Lin, Xi Xuan