AGORA is an item‑response‑theory aggregator that improves the reliability of large language model judges by learning each judge’s sensitivity and rating thresholds from all items jointly, without needing human labels. It places every item on a latent quality scale, allowing judge reliability to be inferred from the data itself. Across three benchmarks, AGORA matches or surpasses the Dawid‑Skene method, especially when evaluation sets are small or judge pools are uneven.
By Junhyung Ahn
The paper argues that the current LLM-as-a-judge evaluation method, which compensates for systematic measurement bias by increasing the number of comparisons, is statistically unsound and computationally wasteful. It identifies that treating LLM judges as neutral ignores documented biases such as position bias, verbosity bias, judge severity, and self‑enhancement. The authors propose a unified latent variable framework that jointly models pairwise and ordinal data while explicitly correcting for these confounders, enabling reliable rankings with far fewer comparisons and negligible additional compute.
By Harshita Katoch, David Antony Selby, Gerrit Gro{\ss}mann, Sebastian Vollmer
arXiv:2607. 08535v1 Announce Type: cross Abstract: An LLM-as-judge score can move even when the candidate responses stay fixed, simply because the evaluator has changed.
By Zongyou Yang, Yinghan Hou, Xiaokun Yang
arXiv:2609.22512v1 Announce Type: new
Abstract: Consensus among LLM judges is often taken as strong evidence that a decision is correct. This assumes that judges make their errors independently. In p...
By Elias Hossain, Niloofar Yousefi, Ser-Nam Lim
arXiv:2601. 21817v2 Announce Type: replace-cross Abstract: Evaluating large language models (LLMs) on open-ended tasks without ground-truth labels is increasingly done via the LLM-as-a-judge paradigm.
By Mingyuan Xu, Xinzi Tan, Jiawei Wu, Doudou Zhou
arXiv:2603.01865v4 Announce Type: replace
Abstract: LLM-as-judge evaluation has become standard practice for open-ended model assessment; however, judges exhibit systematic biases that cannot be aver...
By Ziyi Zhu, Olivier Tieleman, Alexey Bukhtiyarov, Jinghong Chen
arXiv:2605. 02122v2 Announce Type: replace-cross Abstract: Human evaluation remains the primary standard for assessing modern AI systems, yet annotator disagreement, bias, and variability make system rankings fragile under standard majority vote aggregation.
By Akash Bonagiri, Gerard Janno Anderias, Saee Patil, Angelina Lai, Devang Borkar, Gezheng Kang, Ishant Gandhi, Setareh Rafatirad, Houman Homayoun
The paper audits the reliability of large language models used as judges in NLP evaluation. It stresses that verdicts can vary across identical replications, change with presentation order, and that even the most deterministic model only agrees with ground truth 51% of the time. The authors introduce the trustworthy verdict rate (T) to quantify reproducibility, order‑invariance, and accuracy, and show that holistic rubric scoring improves trustworthiness more than any single prompting strategy.
By Vineet Kumar, Darshita Rathore, Anindya Moitra
arXiv:2609.06444v2 Announce Type: cross
Abstract: An LLM judge evaluates outputs at scale. Experts should label only where it is least sure. Its natural escalation signal conflates two uncertainties:...
By Ryan Lail
The paper proposes treating an ensemble of language models as a jury rather than an electorate, where each model’s reliability is inferred from patterns of agreement and disagreement without using labels. Using Expectation Maximisation, the authors estimate each expert’s latent reliability and derive a log‑odds weighted linear scoring rule. On a 1560‑question benchmark, this label‑free approach matches the performance of label‑fitted weights and surpasses the best single model and other label‑free alternatives.
By Alessandro Rozza, Davide Mambelli, Steven Motta, Juri Taresch
The paper introduces a risk‑controlled framework for using large language models (LLMs) as judges in tasks without reference answers. By calibrating uncertainty thresholds on a held‑out set, the method ensures that the false discovery rate of accepted verdicts stays below a user‑specified level α with high probability, using finite‑sample Clopper–Pearson intervals. When the parametric judge lacks confidence, the instance is routed to a retrieval‑augmented mode with a second calibrated threshold, preserving the error guarantee while achieving higher coverage than single‑mode baselines.
arXiv:2606. 13685v1 Announce Type: cross Abstract: LLM-as-a-Judge is now widely used to rank model outputs, train reward models, and populate public leaderboards, but its run-to-run reliability remains under-characterized.
By Abel Yagubyan