Brittlebench: Quantifying LLM robustness via prompt sensitivity

• URL: http://arxiv.org/abs/2603.13285v1
• Date: Fri, 27 Feb 2026 21:12:13 GMT
• Title: Brittlebench: Quantifying LLM robustness via prompt sensitivity
• Abstract summary: We introduce a theoretical framework for quantifying model sensitivity to prompt variants, or brittleness.<n>We design a novel evaluation pipeline, Brittlebench, to holistically evaluate the sensitivity of frontier models.<n>We find that semantics-preserving input perturbations can account for up to half of the performance variance for a given model.
• Score: 44.950999933205985
• License: http://creativecommons.org/licenses/by-sa/4.0/
• Abstract: Existing evaluation methods largely rely on clean, static benchmarks, which can overestimate true model performance by failing to capture the noise and variability inherent in real-world user inputs. This is especially true for language models, which can face human-generated text queries containing mistakes, typos, or alternative ways of phrasing the same question. In this work, we introduce a theoretical framework for quantifying model sensitivity to prompt variants, or brittleness, that can enable us to disentangle data-induced difficulty from prompt-related variability. Using this framework, we design a novel evaluation pipeline, Brittlebench, to holistically evaluate the sensitivity of frontier models. We apply semantics-preserving perturbations to a suite of popular benchmarks, and observe model performance to degrade as much as 12%. However, these perturbations do not affect all models equally: even a single perturbation alters the relative ranking of models in 63% of cases, impacting conclusions about comparative model performance. Decomposing the total variance of both state-of-the-art open-weight and commercial models, we find that semantics-preserving input perturbations can account for up to half of the performance variance for a given model. Brittlebench highlights the need for more robust evaluations and models, and allows us to systematically understand model brittleness.

Related papers

• Valid Best-Model Identification for LLM Evaluation via Low-Rank Factorization [17.34147279018477]
  We propose a principled framework that combines MAB with cheap predicted scores without compromising statistical validity.<n>We derive doubly robust estimators of each model's performance that use the low-rank predictions to reduce variance.<n> Empirical results on real-world benchmarks show that our approach reduces the number of required evaluations, yielding meaningful savings in compute and cost.
  arXiv  Detail & Related papers  (2026-05-11T11:43:28Z)
• A Probabilistic Consensus-Driven Approach for Robust Counterfactual Explanations [2.4788097726838667]
  We propose a novel approach that jointly models the data distribution and the space of plausible model decisions to ensure robustness to model changes.<n>Our method effectively pushes CFEs toward regions that are both plausible and stable across model changes.
  arXiv  Detail & Related papers  (2026-04-19T15:31:18Z)
• Fault-Tolerant Evaluation for Sample-Efficient Model Performance Estimators [13.227055178509524]
  We propose a fault-tolerant evaluation framework that integrates bias and variance considerations within an adjustable tolerance level.<n>We show that proper calibration of $varepsilon$ ensures reliable evaluation across different variance regimes.<n> Experiments on real-world datasets demonstrate that our framework provides comprehensive and actionable insights into estimator behavior.
  arXiv  Detail & Related papers  (2026-02-06T22:14:46Z)
• ConStat: Performance-Based Contamination Detection in Large Language Models [7.305342793164905]
  ConStat is a statistical method that reliably detects and quantifies contamination by comparing performance between a primary and reference benchmark relative to a set of reference models. We demonstrate the effectiveness of ConStat in an extensive evaluation of diverse model architectures, benchmarks, and contamination scenarios.
  arXiv  Detail & Related papers  (2024-05-25T15:36:37Z)
• Semantic Sensitivities and Inconsistent Predictions: Measuring the Fragility of NLI Models [44.56781176879151]
  State-of-the-art Natural Language Inference (NLI) models are sensitive towards minor semantics preserving surface-form variations. We show that semantic sensitivity causes performance degradations of $12.92%$ and $23.71%$ average over $textbfin-$ and $textbfout-of-$ domain settings.
  arXiv  Detail & Related papers  (2024-01-25T14:47:05Z)
• Tailoring Language Generation Models under Total Variation Distance [55.89964205594829]
  The standard paradigm of neural language generation adopts maximum likelihood estimation (MLE) as the optimizing method. We develop practical bounds to apply it to language generation. We introduce the TaiLr objective that balances the tradeoff of estimating TVD.
  arXiv  Detail & Related papers  (2023-02-26T16:32:52Z)
• 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)
• SlimSeg: Slimmable Semantic Segmentation with Boundary Supervision [54.16430358203348]
  We propose a simple but effective slimmable semantic segmentation (SlimSeg) method, which can be executed at different capacities during inference. We show that our proposed SlimSeg with various mainstream networks can produce flexible models that provide dynamic adjustment of computational cost and better performance.
  arXiv  Detail & Related papers  (2022-07-13T14:41:05Z)
• Evaluating the Robustness of Neural Language Models to Input Perturbations [7.064032374579076]
  In this study, we design and implement various types of character-level and word-level perturbation methods to simulate noisy input texts. We investigate the ability of high-performance language models such as BERT, XLNet, RoBERTa, and ELMo in handling different types of input perturbations. The results suggest that language models are sensitive to input perturbations and their performance can decrease even when small changes are introduced.
  arXiv  Detail & Related papers  (2021-08-27T12:31:17Z)
• Firearm Detection via Convolutional Neural Networks: Comparing a Semantic Segmentation Model Against End-to-End Solutions [68.8204255655161]
  Threat detection of weapons and aggressive behavior from live video can be used for rapid detection and prevention of potentially deadly incidents. One way for achieving this is through the use of artificial intelligence and, in particular, machine learning for image analysis. We compare a traditional monolithic end-to-end deep learning model and a previously proposed model based on an ensemble of simpler neural networks detecting fire-weapons via semantic segmentation.
  arXiv  Detail & Related papers  (2020-12-17T15:19:29Z)
• 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)

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.