The paper investigates how many human annotators are equivalent to a panel of 32 large‑language‑model (LLM) judges. By comparing the panel’s label distributions to empirical human labels on three ChaosNLI tasks, the authors find two distinct effective panel sizes: distribution‑error matching yields effective sizes of 2.304, 3.750, and 3.445, while spectral matching gives 4.242, 6.459, and 6.499, indicating a 1.72–1.89× gap. The study also explores how spectral diversity, participation ratio, and panel composition affect effective size, and demonstrates that carefully chosen panels can outperform baseline accuracy while improving effective size.
arXiv:2609.21277v1 Announce Type: cross Abstract: How many human judgments does a panel of language models represent? The answer depends on what is matched. We audit categorical judge panels against...
The study analyzes 373,019 judgments from LLM‑scored benchmarks, decomposing variance into system, item, judge, and interaction components via generalizability theory. It finds that with a single judge, generalizability converges to a ceiling determined by the system‑by‑judge variance, which is substantially lower in pairwise preference settings, allowing one judge to suffice. The research also reveals significant biases in presentation order and highlights that many published win‑rate claims fall below the measured floor of the benchmarks.
arXiv:2607. 28685v1 Announce Type: new Abstract: Agent-safety benchmarks measure different behaviors, and their scores get quoted interchangeably as an agent's safety.
arXiv:2607. 02104v2 Announce Type: replace Abstract: Large language models (LLMs) are increasingly used as cheap, scalable judges that compare candidate outputs pairwise.
arXiv:2608. 06940v1 Announce Type: new Abstract: LLM judge panels are a standard evaluation tool, but prior work reports highly correlated panel errors: nine judges provide roughly the effective information of two independent ones, and aggregation closes only a small fraction of the gap.
arXiv:2608. 15980v1 Announce Type: cross Abstract: Preference benchmarks are built by hiring annotators, and the identity of those annotators is treated as an implementation detail.
arXiv:2608.22432v1 Announce Type: cross Abstract: Multilingual LLM judges produce different evaluator-backbone rankings depending on the prompt language: on an eight-language Agent-as-a-Judge benchma...
arXiv:2606. 24903v2 Announce Type: replace Abstract: Few-shot label acquisition lacks a label-free signal for when additional labels cease to improve accuracy: existing stopping criteria either require a held-out validation set (violating the few-shot premise) or rely on theoretically ungrounded heuristics, so we introduce the spectral saturation index $S(K)=\mathrm{erank}(\hat{\Sigma}_W^{(K)})/K$, the exponential spectral entropy of the pooled within-class covariance normalized by per-class support size $K$, which measures the exploration rate per label and falls below a fixed threshold $\tau=0.
The paper investigates how language‑model judges can make version‑dependent errors when evaluating upgraded agents. Using 35 public coding‑agent submissions, two customer‑service agents, and over a thousand expert‑labeled trajectories, the authors show that fixed judges often reject task‑conditioned error invariance and can incorrectly approve failed patches, especially as agent capability increases. Paired audits of current outputs reduce interval width only marginally, and the study concludes that independent human patch review is still necessary.
The article examines how classical test theory reliability statistics misrepresent the performance of large language model (LLM) judges. It shows that internal‑consistency coefficients, the dependability index, and Livingston‑Lewis accuracy each conflate judge error with item design or criterion validity, making it impossible to attribute a single reliability value to the judge alone. The authors argue that such misattribution can influence deployment decisions and documentation.
arXiv:2607. 17136v1 Announce Type: cross Abstract: Agentic computer-use RL is reported in single runs, and those numbers mislead.
The paper introduces a pluralistic agreement index, Gamma, to quantify how often wrong runs of large language models (LLMs) agree with the majority consensus. By decomposing Gamma into a mechanical component and a preference‑unexplained residual, the authors show that on GPT‑4.1 the mechanical part explains most of the agreement on multiple‑choice benchmarks but only about half on open‑domain tasks, revealing a residual bias that can cause self‑consistency to backfire on hard questions. The study provides a quantitative framework for understanding when majority voting over LLM samples improves or harms accuracy, without proposing new voting methods.