Stress-Testing Model Specs Reveals Character Differences among Language Models

• URL: http://arxiv.org/abs/2510.07686v2
• Date: Thu, 23 Oct 2025 07:31:33 GMT
• Title: Stress-Testing Model Specs Reveals Character Differences among Language Models
• Authors: Jifan Zhang, Henry Sleight, Andi Peng, John Schulman, Esin Durmus,
• Abstract summary: Large language models (LLMs) are increasingly trained from AI constitutions and model specifications.<n>We present a systematic methodology for stress-testing model character specifications.<n>We identify numerous cases of principle contradictions and interpretive ambiguities in current model specs.
• Score: 23.505192393830807
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Large language models (LLMs) are increasingly trained from AI constitutions and model specifications that establish behavioral guidelines and ethical principles. However, these specifications face critical challenges, including internal conflicts between principles and insufficient coverage of nuanced scenarios. We present a systematic methodology for stress-testing model character specifications, automatically identifying numerous cases of principle contradictions and interpretive ambiguities in current model specs. We stress test current model specs by generating scenarios that force explicit tradeoffs between competing value-based principles. Using a comprehensive taxonomy we generate diverse value tradeoff scenarios where models must choose between pairs of legitimate principles that cannot be simultaneously satisfied. We evaluate responses from twelve frontier LLMs across major providers (Anthropic, OpenAI, Google, xAI) and measure behavioral disagreement through value classification scores. Among these scenarios, we identify over 70,000 cases exhibiting significant behavioral divergence. Empirically, we show this high divergence in model behavior strongly predicts underlying problems in model specifications. Through qualitative analysis, we provide numerous example issues in current model specs such as direct contradiction and interpretive ambiguities of several principles. Additionally, our generated dataset also reveals both clear misalignment cases and false-positive refusals across all of the frontier models we study. Lastly, we also provide value prioritization patterns and differences of these models.

Related papers

• Improving Group Robustness on Spurious Correlation via Evidential Alignment [26.544938760265136]
  Deep neural networks often learn and rely on spurious correlations, i.e., superficial associations between non-causal features and the targets.<n>Existing methods typically mitigate this issue by using external group annotations or auxiliary deterministic models.<n>We propose Evidential Alignment, a novel framework that leverages uncertainty quantification to understand the behavior of the biased models.
  arXiv  Detail & Related papers  (2025-06-12T22:47:21Z)
• Delphos: A reinforcement learning framework for assisting discrete choice model specification [0.0]
  We introduce Delphos, a deep reinforcement learning framework for assisting the discrete choice model specification process.<n>In this setting, an agent learns to specify well-performing model candidates by choosing a sequence of modelling actions.<n>We evaluate Delphos on both simulated and empirical datasets, varying the size of the modelling space and the reward function.
  arXiv  Detail & Related papers  (2025-06-06T15:40:16Z)
• Preference Learning for AI Alignment: a Causal Perspective [55.2480439325792]
  We frame this problem in a causal paradigm, providing the rich toolbox of causality to identify persistent challenges.<n>Inheriting from the literature of causal inference, we identify key assumptions necessary for reliable generalisation.<n>We illustrate failure modes of naive reward models and demonstrate how causally-inspired approaches can improve model robustness.
  arXiv  Detail & Related papers  (2025-06-06T10:45:42Z)
• Relative Overfitting and Accept-Reject Framework [5.465098504510676]
  We propose an ensemble framework that governs how models are segmented to ensure performance improvement.<n>We detail the patterns of this framework within the domain of NLP and briefly describe its to other fields, such as computer vision (CV) and AI for science.
  arXiv  Detail & Related papers  (2025-05-12T17:36:14Z)
• On the Reasoning Capacity of AI Models and How to Quantify It [0.0]
  Large Language Models (LLMs) have intensified the debate surrounding the fundamental nature of their reasoning capabilities.<n>While achieving high performance on benchmarks such as GPQA and MMLU, these models exhibit limitations in more complex reasoning tasks.<n>We propose a novel phenomenological approach that goes beyond traditional accuracy metrics to probe the underlying mechanisms of model behavior.
  arXiv  Detail & Related papers  (2025-01-23T16:58:18Z)
• Predictive Churn with the Set of Good Models [61.00058053669447]
  This paper explores connections between two seemingly unrelated concepts of predictive inconsistency.<n>The first, known as predictive multiplicity, occurs when models that perform similarly produce conflicting predictions for individual samples.<n>The second concept, predictive churn, examines the differences in individual predictions before and after model updates.
  arXiv  Detail & Related papers  (2024-02-12T16:15:25Z)
• Challenges to Evaluating the Generalization of Coreference Resolution Models: A Measurement Modeling Perspective [69.50044040291847]
  We show how multi-dataset evaluations risk conflating different factors concerning what, precisely, is being measured. This makes it difficult to draw more generalizable conclusions from these evaluations.
  arXiv  Detail & Related papers  (2023-03-16T05:32:02Z)
• Are Neural Topic Models Broken? [81.15470302729638]
  We study the relationship between automated and human evaluation of topic models. We find that neural topic models fare worse in both respects compared to an established classical method.
  arXiv  Detail & Related papers  (2022-10-28T14:38:50Z)
• Enhancing Model Robustness and Fairness with Causality: A Regularization Approach [15.981724441808147]
  Recent work has raised concerns on the risk of spurious correlations and unintended biases in machine learning models. We propose a simple and intuitive regularization approach to integrate causal knowledge during model training. We build a predictive model that relies more on causal features and less on non-causal features.
  arXiv  Detail & Related papers  (2021-10-03T02:49:33Z)
• Characterizing Fairness Over the Set of Good Models Under Selective Labels [69.64662540443162]
  We develop a framework for characterizing predictive fairness properties over the set of models that deliver similar overall performance. We provide tractable algorithms to compute the range of attainable group-level predictive disparities. We extend our framework to address the empirically relevant challenge of selectively labelled data.
  arXiv  Detail & Related papers  (2021-01-02T02:11:37Z)
• Improving the Reconstruction of Disentangled Representation Learners via Multi-Stage Modeling [54.94763543386523]
  Current autoencoder-based disentangled representation learning methods achieve disentanglement by penalizing the ( aggregate) posterior to encourage statistical independence of the latent factors. We present a novel multi-stage modeling approach where the disentangled factors are first learned using a penalty-based disentangled representation learning method. Then, the low-quality reconstruction is improved with another deep generative model that is trained to model the missing correlated latent variables.
  arXiv  Detail & Related papers  (2020-10-25T18:51:15Z)
• Accounting for Unobserved Confounding in Domain Generalization [107.0464488046289]
  This paper investigates the problem of learning robust, generalizable prediction models from a combination of datasets. Part of the challenge of learning robust models lies in the influence of unobserved confounders. We demonstrate the empirical performance of our approach on healthcare data from different modalities.
  arXiv  Detail & Related papers  (2020-07-21T08:18:06Z)

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.