CVPR Findings2026
Zero-Shot Textual Explanations via Translating Decision-Critical Features
Toshinori Yamauchi, Hiroshi Kera, Kazuhiko Kawamoto
Abstract
We propose TEXTER, which overcomes this limitation by isolating decision-critical features before alignment. TEXTER identifies the neurons contributing to the prediction and emphasizes the features encoded in those neurons-i.e., the decision-critical features. It then maps these emphasized features into the CLIP feature space to retrieve textual explanations that reflect the model's reasoning. A sparse autoencoder further improves interpretability, particularly for Transformer architectures. Extensive experiments show that TEXTER provides more faithful and interpretable explanations than existing methods. The code is available at https://github.com/tttt-0814/ TEXTER.
Ask Lune
Know this paper in minutes, not an afternoon.
Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.
Lineage
- Emerging Properties in Self-Supervised Vision Transformers cites An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
- Visual Instruction Tuning cites High-Resolution Image Synthesis with Latent Diffusion Models
- Visual Instruction Tuning cites Learning Transferable Visual Models From Natural Language Supervision
- Learning Transferable Visual Models From Natural Language Supervision cites An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
Builds on33
- Learning Transferable Visual Models From Natural Language SupervisionAlec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh et al.ICML 2021 · 47,906 citations
- An Image is Worth 16x16 Words: Transformers for Image Recognition at ScaleAlexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn et al.ICLR 2021 · 21,477 citations
- High-Resolution Image Synthesis with Latent Diffusion ModelsRobin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser et al.CVPR 2022 · 13,123 citations
- Visual Instruction TuningHaotian Liu, Chunyuan Li, Qingyang Wu, Yong Jae LeeNeurIPS 2023 · 11,349 citations
- Emerging Properties in Self-Supervised Vision TransformersMathilde Caron, Hugo Touvron, Ishan Misra, Hervé Jégou et al.ICCV 2021 · 8,921 citations
Related papers
- Graph-Guided Textual Explanation Generation FrameworkShuzhou Yuan, Jingyi Sun, Ran Zhang, Michael Färber et al.EMNLP 2025 · 1 citation
- Class Activation Values: Lucid and Faithful Visual Interpretations for CLIP-based Text-Image RetrievalsPengxu Chen, Huazhong Liu, Jihong Ding, Xinghao Huang et al.SIGIR 2025 · 5 citations
- Interpreting CLIP with Sparse Linear Concept Embeddings (SpLiCE)Usha Bhalla, Alex Oesterling, Suraj Srinivas, Flávio P. Calmon et al.NeurIPS 2024 · 146 citations
- VL-SAE: Interpreting and Enhancing Vision-Language Alignment with a Unified Concept SetShufan Shen, Junshu Sun, Qingming Huang, Shuhui WangNeurIPS 2025 · 13 citations
- MICLIP: Learning to Interpret Representation in Vision ModelsYingdong Shi, Zhiyu Yang, Changming Li, Jingyi Yu et al.ICLR 2026
