# SUB: Benchmarking CBM Generalization via Synthetic Attribute Substitutions

arXiv: https://arxiv.org/abs/2507.23784 · HTML: https://mlsanity.com/paper/2507.23784/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- Builds on 60 of its 71 references that are in the map.
- First posted on arXiv 31 Jul 2025.

## Builds on (most cited first, 30 of 60)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [High-Resolution Image Synthesis with Latent Diffusion Models](https://mlsanity.com/paper/2112.10752/) (2021)
- [Classifier-Free Diffusion Guidance](https://mlsanity.com/paper/2207.12598/) (2022)
- [Adding Conditional Control to Text-to-Image Diffusion Models](https://mlsanity.com/paper/2302.05543/) (2023)
- [SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesis](https://mlsanity.com/paper/2307.01952/) (2023)
- [Sigmoid Loss for Language Image Pre-Training](https://mlsanity.com/paper/2303.15343/) (2023)
- [InternVL: Scaling up Vision Foundation Models and Aligning for Generic Visual-Linguistic Tasks](https://mlsanity.com/paper/2312.14238/) (2023)
- [Expanding Performance Boundaries of Open-Source Multimodal Models with Model, Data, and Test-Time Scaling](https://mlsanity.com/paper/2412.05271/) (2024)
- [CLEVR: A Diagnostic Dataset for Compositional Language and Elementary Visual Reasoning](https://mlsanity.com/paper/1612.06890/) (2016)
- [Sparse Autoencoders Find Highly Interpretable Features in Language Models](https://mlsanity.com/paper/2309.08600/) (2023)
- [SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense Features](https://mlsanity.com/paper/2502.14786/) (2025)
- [FlowNet: Learning Optical Flow with Convolutional Networks](https://mlsanity.com/paper/1504.06852/) (2015)
- [How Far Are We to GPT-4V? Closing the Gap to Commercial Multimodal Models with Open-Source Suites](https://mlsanity.com/paper/2404.16821/) (2024)
- [Concept Bottleneck Models](https://mlsanity.com/paper/2007.04612/) (2020)
- [EVA-CLIP: Improved Training Techniques for CLIP at Scale](https://mlsanity.com/paper/2303.15389/) (2023)
- [This Looks Like That: Deep Learning for Interpretable Image Recognition](https://mlsanity.com/paper/1806.10574/) (2018)
- [Compositional Visual Generation with Composable Diffusion Models](https://mlsanity.com/paper/2206.01714/) (2022)
- [Towards Robust Interpretability with Self-Explaining Neural Networks](https://mlsanity.com/paper/1806.07538/) (2018)
- [Uni-ControlNet: All-in-One Control to Text-to-Image Diffusion Models](https://mlsanity.com/paper/2305.16322/) (2023)
- [Visual Classification via Description from Large Language Models](https://mlsanity.com/paper/2210.07183/) (2022)
- [Is synthetic data from generative models ready for image recognition?](https://mlsanity.com/paper/2210.07574/) (2022)
- [Language in a Bottle: Language Model Guided Concept Bottlenecks for Interpretable Image Classification](https://mlsanity.com/paper/2211.11158/) (2022)
- [Label-Free Concept Bottleneck Models](https://mlsanity.com/paper/2304.06129/) (2023)
- [Reduce, Reuse, Recycle: Compositional Generation with Energy-Based Diffusion Models and MCMC](https://mlsanity.com/paper/2302.11552/) (2023)
- [Fake it till you make it: Learning transferable representations from synthetic ImageNet clones](https://mlsanity.com/paper/2212.08420/) (2022)
- [Promises and Pitfalls of Black-Box Concept Learning Models](https://mlsanity.com/paper/2106.13314/) (2021)
- [ReNO: Enhancing One-step Text-to-Image Models through Reward-based Noise Optimization](https://mlsanity.com/paper/2406.04312/) (2024)
- [Do Concept Bottleneck Models Learn as Intended?](https://mlsanity.com/paper/2105.04289/) (2021)
- [Diversify Your Vision Datasets with Automatic Diffusion-Based Augmentation](https://mlsanity.com/paper/2305.16289/) (2023)
- [Interpretable Image Classification with Differentiable Prototypes Assignment](https://mlsanity.com/paper/2112.02902/) (2021)

## Built on by (most cited first, 6 of 6)

- [Stitch: Training-Free Position Control in Multimodal Diffusion Transformers](https://mlsanity.com/paper/2509.26644/) (2025)
- [Evaluating the Interpretability of Sparse Autoencoders with Concept Annotations](https://mlsanity.com/paper/2606.24716/) (2026)
- [Seeing Through Circuits: Faithful Mechanistic Interpretability for Vision Transformers](https://mlsanity.com/paper/2604.14477/) (2026)
- [Logit Distance Bounds Representational Similarity](https://mlsanity.com/paper/2602.15438/) (2026)
- [The Latent Color Subspace: Emergent Order in High-Dimensional Chaos](https://mlsanity.com/paper/2603.12261/) (2026)
- [Metonymy in vision models undermines attention-based interpretability](https://mlsanity.com/paper/2605.06095/) (2026)

## Abstract (arXiv)

Concept Bottleneck Models (CBMs) and other concept-based interpretable models show great promise for making AI applications more transparent, which is essential in fields like medicine. Despite their success, we demonstrate that CBMs struggle to reliably identify the correct concepts under distribution shifts. To assess the robustness of CBMs to concept variations, we introduce SUB: a fine-grained image and concept benchmark containing 38,400 synthetic images based on the CUB dataset. To create SUB, we select a CUB subset of 33 bird classes and 45 concepts to generate images which substitute a specific concept, such as wing color or belly pattern. We introduce a novel Tied Diffusion Guidance (TDG) method to precisely control generated images, where noise sharing for two parallel denoising processes ensures that both the correct bird class and the correct attribute are generated. This novel benchmark enables rigorous evaluation of CBMs and similar interpretable models, contributing to the development of more robust methods. Our code is available at https://github.com/ExplainableML/sub and the dataset at http://huggingface.co/datasets/Jessica-bader/SUB.