Deployment-Relevant Alignment Cannot Be Inferred from Model-Level Evaluation Alone

• URL: http://arxiv.org/abs/2605.04454v1
• Date: Wed, 06 May 2026 03:28:30 GMT
• Title: Deployment-Relevant Alignment Cannot Be Inferred from Model-Level Evaluation Alone
• Abstract summary: Alignment evaluation in machine learning has largely become evaluation of models.<n>This paper argues that deployment-relevant alignment cannot be inferred from model-level evaluation alone.
• Score: 11.663456969895462
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Alignment evaluation in machine learning has largely become evaluation of models. Influential benchmarks score model outputs under fixed inputs, such as truthfulness, instruction following, or pairwise preference, and these scores are often used to support claims about deployed alignment. This paper argues that deployment-relevant alignment cannot be inferred from model-level evaluation alone. Alignment claims should instead be indexed to the level at which evidence is collected: model-level, response-level, interaction-level, or deployment-level. Two studies support this position. First, a structured audit of eleven alignment benchmarks, extended to a sixteen-benchmark corpus, dual-coded against an eight-dimension rubric with Cohen's kappa = 0.87, finds that user-facing verification support is absent across every benchmark examined, while process steerability is nearly absent. The few interactional benchmarks identified, including tau-bench, CURATe, Rifts, and Common Ground, remain fragmented in coverage, and benchmark construction rather than data source determines what is measured. Second, a blinded cross-model stress test using 180 transcripts across three frontier models and four scaffolds finds that the same verification scaffold raises one model's verification support to ceiling while leaving another categorically unchanged. This shows that scaffold efficacy is model-dependent and that the gap identified by the audit cannot be closed at the model level alone. We propose a system-level evaluation agenda: alignment profiles instead of single scores, fixed-scaffolding protocols for comparable interactional evaluation, and reporting templates that make the inferential distance between evaluation evidence and deployment claims explicit.

Related papers

• The Double Measurement Confound in Agent Benchmarks: De-Scaffolding, Ground-Truth Scoring, and Reliability Beyond the Mean [2.822987192839875]
  Agent benchmarks are increasingly used to compare large language models (LLMs) and guide deployment decisions.<n>We identify a double measurement confound: execution-critical decisions are performed by a fixed scaffold instead of the model, while the scorer evaluates outputs using criteria that may not reflect task correctness.<n>We unify these issues within a measurement-theoretic framework that characterizes when benchmark scores can be interpreted as evidence of model capability.
  arXiv  Detail & Related papers  (2026-09-06T21:23:11Z)
• Beyond Polarization: The Generative Constraint of Chain-of-Thought in Pointwise Reranking [28.78994214839467]
  In pointwise document reranking, Chain-of-Thought models typically underperform direct scoring models.<n>Our empirical study first confirms that this gap is stable across scales up to 32B parameters, ruling out model and data capacity confounders.<n>We then apply stress tests utilizing reinforcement learning, fine-grained supervision, and architectural decoupling to explicitly repair these deviations.
  arXiv  Detail & Related papers  (2026-08-31T07:51:49Z)
• What AI Red-Team Evaluations Can and Cannot Prove [0.0]
  Red-team evaluations of AI models support some claims and not others, and the boundary between the two is calculable.<n>We define the evidential ceiling of an evaluation as the largest factor by which one result can move belief under a fixed testing budget.<n>We find that above a calculable harm rate, a benchmark of modest size certifies a category to a stated evidentiary standard.
  arXiv  Detail & Related papers  (2026-07-23T18:34:12Z)
• PerceptionRubrics: Calibrating Multimodal Evaluation to Human Perception [75.70273182182397]
  PerceptionRubrics is a rubric-based evaluation framework.<n>It addresses the gap between saturated benchmark scores and real-world brittleness.<n>It pairs 1,038 information-dense images with over 10,000 instance-specific rubrics.
  arXiv  Detail & Related papers  (2026-06-26T17:59:15Z)
• DualEval: Joint Model-Item Calibration for Unified LLM Evaluation [67.99119926012686]
  DualEval is a latent model-item calibration framework that represents models and evaluation items in a shared space.<n>We apply DualEval across four domains: coding, math, miscellaneous domain-knowledge tasks, and generic everyday user queries.
  arXiv  Detail & Related papers  (2026-06-24T22:40:46Z)
• What Accuracy and Gradient Cosine Miss: Evaluating Feedback Alignment via Scale Stability, Reference Validity, and Depth Utility [48.10132234701036]
  We propose a diagnostic evaluation protocol based on three checks -- scale stability, reference validity, and depth utility -- together with per-layer rather than aggregate cosine reporting.<n>Across multiple architectures and methods, our protocol identifies all failures with wide calibration margins.
  arXiv  Detail & Related papers  (2026-06-19T06:04:17Z)
• Beyond Static Leaderboards: Predictive Validity for the Evaluation of LLM Agents [19.290397279235084]
  This paper aggregates the largest coordinated deep-dive of one MCP-based industrial-agent benchmark to date.<n>We argue that aggregate-score leaderboards systematically underspecify deployed-agent evaluation.
  arXiv  Detail & Related papers  (2026-06-18T02:02:46Z)
• Interactive Evaluation Requires a Design Science [90.62695599188204]
  Large language models (LLMs) are increasingly deployed as systems that act over time through tools, environments, users, and other agents.<n>The field has begun to build interactive benchmarks, but the resulting landscape is fragmented.<n>This position paper argues that interactive evaluation should be treated as a principled evaluation paradigm.
  arXiv  Detail & Related papers  (2026-05-18T04:03:18Z)
• When Verification Fails: How Compositionally Infeasible Claims Escape Rejection [30.404615085122046]
  We show that existing verification benchmarks cannot distinguish between rigorous claim verification and simple salient-constraint reliance.<n>We construct compositionally infeasible claims where the salient constraint is supported but a non-salient constraint is contradicted.<n>We show that different models occupy distinct positions on a shared ROC curve, indicating that the gap between model families reflects differences in verification threshold rather than underlying reasoning ability.
  arXiv  Detail & Related papers  (2026-04-13T04:48:20Z)
• K$α$LOS finds Consensus: A Meta-Algorithm for Evaluating Inter-Annotator Agreement in Complex Vision Tasks [4.297070083645049]
  We propose K$$LOS, a unified meta-algorithm that generalizes the "Localization First" principle to dataset quality evaluation.<n>By resolving spatial correspondence before assessing agreement, our framework transforms complex-categorical problems into nominal reliability.
  arXiv  Detail & Related papers  (2026-03-28T08:54:05Z)
• Learning More from Less: Unlocking Internal Representations for Benchmark Compression [37.69575776639016]
  We introduce REPCORE, which aligns heterogeneous hidden states into a unified latent space to construct representative coresets.<n>Experiments on five benchmarks and over 200 models show consistent gains over output-based baselines in ranking correlation and estimation accuracy.
  arXiv  Detail & Related papers  (2026-01-31T13:11:39Z)
• SpatialBench-UC: Uncertainty-Aware Evaluation of Spatial Prompt Following in Text-to-Image Generation [0.0]
  SpatialBench-UC is a small, reproducible benchmark for pairwise spatial relations.<n>We release a benchmark package, versioned prompts, pinned configs, per-sample checker outputs, and report tables.<n>We evaluate three baselines, Stable Diffusion 1.5, SD 1.5 BoxDiff, and SD 1.4 GLIGEN.
  arXiv  Detail & Related papers  (2026-01-19T23:37:10Z)
• PaperAudit-Bench: Benchmarking Error Detection in Research Papers for Critical Automated Peer Review [54.141490756509306]
  We introduce PaperAudit-Bench, which consists of two components: PaperAudit-Dataset, an error dataset, and PaperAudit-Review, an automated review framework.<n>Experiments on PaperAudit-Bench reveal large variability in error detectability across models and detection depths.<n>We show that the dataset supports training lightweight LLM detectors via SFT and RL, enabling effective error detection at reduced computational cost.
  arXiv  Detail & Related papers  (2026-01-07T04:26:12Z)
• 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)
• SCORE: A Semantic Evaluation Framework for Generative Document Parsing [2.5101597298392098]
  Multi-modal generative document parsing systems produce semantically correct yet structurally divergent outputs.<n>Conventional metrics-CER, WER, IoU, or TEDS-misclassify such diversity as error, penalizing valid interpretations and obscuring system behavior.<n>We introduce SCORE, an interpretation-agnostic framework that integrates (i) adjusted edit distance for robust content fidelity, (ii) token-level diagnostics to distinguish hallucinations from omissions, (iii) table evaluation with spatial tolerance and semantic alignment, and (iv) hierarchy-aware consistency checks.
  arXiv  Detail & Related papers  (2025-09-16T16:06:19Z)
• Appeal: Allow Mislabeled Samples the Chance to be Rectified in Partial Label Learning [55.4510979153023]
  In partial label learning (PLL), each instance is associated with a set of candidate labels among which only one is ground-truth. To help these mislabeled samples "appeal," we propose the first appeal-based framework.
  arXiv  Detail & Related papers  (2023-12-18T09:09:52Z)

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.