# Beyond Benchmarks: Using VLMs to Reveal Systematic Classification Failures Under Real World Conditions

arXiv: https://arxiv.org/abs/2609.11126 · HTML: https://mlsanity.com/paper/2609.11126/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 18 of its 33 references that are in the map.
- First posted on arXiv 10 Sep 2026.

## Builds on (most cited first, 18 of 18)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localization](https://mlsanity.com/paper/1610.02391/) (2016)
- [UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction](https://mlsanity.com/paper/1802.03426/) (2018)
- [Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization](https://mlsanity.com/paper/1911.08731/) (2019)
- [OpenAI GPT-5 System Card](https://mlsanity.com/paper/2601.03267/) (2025)
- [Towards Out-Of-Distribution Generalization: A Survey](https://mlsanity.com/paper/2108.13624/) (2021)
- [Perception Encoder: The best visual embeddings are not at the output of the network](https://mlsanity.com/paper/2504.13181/) (2025)
- [Domino: Discovering Systematic Errors with Cross-Modal Embeddings](https://mlsanity.com/paper/2203.14960/) (2022)
- [ZeroCap: Zero-Shot Image-to-Text Generation for Visual-Semantic Arithmetic](https://mlsanity.com/paper/2111.14447/) (2021)
- [PerceptionLM: Open-Access Data and Models for Detailed Visual Understanding](https://mlsanity.com/paper/2504.13180/) (2025)
- [Distilling Model Failures as Directions in Latent Space](https://mlsanity.com/paper/2206.14754/) (2022)
- [YOLO26: Key Architectural Enhancements and Performance Benchmarking for Real-Time Object Detection](https://mlsanity.com/paper/2509.25164/) (2025)
- [Interpreting the linear structure of vision-language model embedding spaces](https://mlsanity.com/paper/2504.11695/) (2025)
- [Robustness of Object Detectors in Degrading Weather Conditions](https://mlsanity.com/paper/2106.08795/) (2021)
- [FACTS: First Amplify Correlations and Then Slice to Discover Bias](https://mlsanity.com/paper/2309.17430/) (2023)
- [HiBug2: Efficient and Interpretable Error Slice Discovery for Comprehensive Model Debugging](https://mlsanity.com/paper/2501.16751/) (2025)
- [Steerable Visual Representations](https://mlsanity.com/paper/2604.02327/) (2026)
- [Occlusion Robustness of CLIP for Military Vehicle Classification](https://mlsanity.com/paper/2508.20760/) (2025)

## Abstract (arXiv)

Verification and validation (V&V) of classification models is crucial to enable a wide range of sensor processing applications. Currently, the V&V process relies on time-consuming manual inspection of erroneous samples to find meaningful patterns. This work explores the use of Vision Language Models (VLMs) to speed up this laborious process. VLMs are trained to embed images into a semantically meaningful vector representation, from which human-interpretable systematic errors can be distilled. Deploying such VLM-based methods in a defence context introduces two major challenges: (1) the defence domain is underrepresented in the training data of VLMs, and (2) surroundings and context are less diverse than for other domains. This study provides an initial assessment of the suitability of VLM-based methods for V&V of defence applications. We propose a VLM-based error slice detection (ESD) method that independently groups and labels systematic errors made by a classification model. We demonstrate that this method is able to identify operationally-relevant artificially added perturbations in a non-military dataset. In a military context, our method clusters and describes images based on their surroundings, but also exhibits overlap between cluster descriptions. We further investigate the difference in embedding variation between our military and non-military dataset, which remains a topic of interest. Although the results do not yet warrant fully automated V&V through VLM-based ESD, they show that VLMs could be used to accelerate V&V processes in the future.