Distribution-Calibrated Inference time compute for Thinking LLM-as-a-Judge

• URL: http://arxiv.org/abs/2512.03019v1
• Date: Tue, 02 Dec 2025 18:46:47 GMT
• Title: Distribution-Calibrated Inference time compute for Thinking LLM-as-a-Judge
• Authors: Hamid Dadkhahi, Firas Trabelsi, Parker Riley, Juraj Juraska, Mehdi Mirzazadeh,
• Abstract summary: Thinking Large Language Models (LLMs) used as judges for pairwise preferences remain noisy at the single-sample level.<n>We study inference-time compute (ITC) for evaluators that generate n independent thinking-rating samples per item.
• Score: 5.855996386998925
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: Thinking Large Language Models (LLMs) used as judges for pairwise preferences remain noisy at the single-sample level, and common aggregation rules (majority vote, soft self-consistency, or instruction-based self-aggregation) are inconsistent when ties are allowed. We study inference-time compute (ITC) for evaluators that generate n independent thinking-rating samples per item, and propose a principled, distribution-calibrated aggregation scheme. Our method models three-way preferences with a Bradley-Terry-Davidson formulation on rating counts, leveraging both polarity (margin among non-ties) and decisiveness (non-tie rate) to distinguish narrow margins from strong consensus. Across various evaluation benchmarks, our approach consistently reduces MAE and increases pairwise accuracy versus standard baselines, and when evaluated against human-consensus meta-labels, matches or exceeds individual human raters. These results show that carefully allocating ITC and aggregating with distribution-aware methods turns noisy individual model judgments into reliable ratings for evaluation.

Related papers

• Who can we trust? LLM-as-a-jury for Comparative Assessment [42.32900791516691]
  Large language models (LLMs) are increasingly applied as automatic evaluators for natural language generation assessment.<n>LLMs judges vary substantially in performance across tasks and aspects, and their judgment probabilities may be biased and inconsistent.<n>We propose BT-sigma, a judge-aware extension of the Bradley-Terry model that introduces a discriminator parameter for each judge to jointly infer item rankings and judge reliability from pairwise comparisons alone.
  arXiv  Detail & Related papers  (2026-02-18T17:04:02Z)
• FairJudge: An Adaptive, Debiased, and Consistent LLM-as-a-Judge [10.584937371987742]
  Existing LLM-as-a-Judge systems suffer from limited adaptivity to task- and domain-specific evaluation criteria.<n>We propose FairJudge, an adaptive, debiased, and consistent LLM-as-a-Judge.
  arXiv  Detail & Related papers  (2026-02-06T11:35:32Z)
• Dependence-Aware Label Aggregation for LLM-as-a-Judge via Ising Models [55.94503936470247]
  Large-scale AI evaluation increasingly relies on aggregating binary judgments from $K$ annotators, including judges.<n>Most classical methods assume annotators are conditionally independent given the true label $Yin0,1$, an assumption often violated by LLM judges.<n>We study label aggregation through a hierarchy of dependence-aware models based on Ising graphical models and latent factors.
  arXiv  Detail & Related papers  (2026-01-29T21:26:50Z)
• A Judge-Aware Ranking Framework for Evaluating Large Language Models without Ground Truth [4.9467757325435775]
  evaluating large language models (LLMs) on open-ended tasks is increasingly done via the LLM-as-a-judge paradigm.<n>Treating all judges equally can yield biased leaderboards and misleading uncertainty estimates.<n>We propose a judge-aware ranking framework that extends the Bradley-Terry-Luce model by introducing judge-specific discrimination parameters.
  arXiv  Detail & Related papers  (2026-01-29T15:01:28Z)
• Reference-Free Rating of LLM Responses via Latent Information [53.463883683503106]
  We study the common practice of asking a judge model to assign Likert-scale scores to free-text responses.<n>We then propose and evaluate Latent Judges, which derive scalar ratings from internal model signals.<n>Across a broad suite of pairwise and single-rating benchmarks, latent methods match or surpass standard prompting.
  arXiv  Detail & Related papers  (2025-09-29T12:15:52Z)
• TrustJudge: Inconsistencies of LLM-as-a-Judge and How to Alleviate Them [58.04324690859212]
  Large Language Models (LLMs) as automated evaluators (LLM-as-a-judge) has revealed critical inconsistencies in current evaluation frameworks.<n>We identify two fundamental types of inconsistencies: Score-Comparison Inconsistency and Pairwise Transitivity Inconsistency.<n>We propose TrustJudge, a probabilistic framework that addresses these limitations through two key innovations.
  arXiv  Detail & Related papers  (2025-09-25T13:04:29Z)
• CCE: Confidence-Consistency Evaluation for Time Series Anomaly Detection [56.302586730134806]
  We introduce Confidence-Consistency Evaluation (CCE), a novel evaluation metric.<n>CCE simultaneously measures prediction confidence and uncertainty consistency.<n>We also establish RankEval, a benchmark for comparing the ranking capabilities of various metrics.
  arXiv  Detail & Related papers  (2025-09-01T03:38:38Z)
• Pairwise or Pointwise? Evaluating Feedback Protocols for Bias in LLM-Based Evaluation [57.380464382910375]
  We show that the choice of feedback protocol for evaluation can significantly affect evaluation reliability and induce systematic biases.<n>We find that generator models can flip preferences by embedding distractor features.<n>We offer recommendations for choosing feedback protocols based on dataset characteristics and evaluation objectives.
  arXiv  Detail & Related papers  (2025-04-20T19:05:59Z)
• Compare without Despair: Reliable Preference Evaluation with Generation Separability [20.50638483427141]
  We introduce a measure, separability, which estimates how suitable a test instance is for pairwise preference evaluation. For a candidate test instance, separability samples multiple generations from a pair of models, and measures how distinguishable the two sets of generations are. Experiments show that instances with high separability values yield more consistent preference ratings from both human- and auto-raters.
  arXiv  Detail & Related papers  (2024-07-02T01:37:56Z)
• A Statistical Analysis of Summarization Evaluation Metrics using Resampling Methods [60.04142561088524]
  We find that the confidence intervals are rather wide, demonstrating high uncertainty in how reliable automatic metrics truly are. Although many metrics fail to show statistical improvements over ROUGE, two recent works, QAEval and BERTScore, do in some evaluation settings.
  arXiv  Detail & Related papers  (2021-03-31T18:28:14Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.