eXCV Workshop at ECCV 2026 https://excv-workshop.github.io/ eXCV Workshop at ECCV 2026 Source Themes Academic (https://sourcethemes.com/academic/)en-us© 2026 eXCV Workshop Organizers https://excv-workshop.github.io/img/icon.png eXCV Workshop at ECCV 2026 https://excv-workshop.github.io/ Ask Twice, Look Twice: Prompt Echoing Resolves the Question-First Paradox in Vision-Language Models https://excv-workshop.github.io/publication/ask-twice-look-twice-prompt-echoing-resolves-the-question-first-paradox-in-vision-language-models/ Mon, 01 Jan 0001 00:00:00 +0000 https://excv-workshop.github.io/publication/ask-twice-look-twice-prompt-echoing-resolves-the-question-first-paradox-in-vision-language-models/ Attention (as Discrete-Time Markov) Chains https://excv-workshop.github.io/publication/attention-as-discrete-time-markov-chains/ Mon, 01 Jan 0001 00:00:00 +0000 https://excv-workshop.github.io/publication/attention-as-discrete-time-markov-chains/ CBX-Bench: A Human-Aligned MLLM Council for Benchmarking Concept Bottleneck Model Explanations https://excv-workshop.github.io/publication/cbx-bench-a-human-aligned-mllm-council-for-benchmarking-concept-bottleneck-model-explanations/ Mon, 01 Jan 0001 00:00:00 +0000 https://excv-workshop.github.io/publication/cbx-bench-a-human-aligned-mllm-council-for-benchmarking-concept-bottleneck-model-explanations/ CFM: Language-aligned Concept Foundation Model for Vision https://excv-workshop.github.io/publication/cfm-language-aligned-concept-foundation-model-for-vision/ Mon, 01 Jan 0001 00:00:00 +0000 https://excv-workshop.github.io/publication/cfm-language-aligned-concept-foundation-model-for-vision/ DiMaS: Distribution Matching for Steering Vision-Language-Action Models https://excv-workshop.github.io/publication/dimas-distribution-matching-for-steering-vision-language-action-models/ Mon, 01 Jan 0001 00:00:00 +0000 https://excv-workshop.github.io/publication/dimas-distribution-matching-for-steering-vision-language-action-models/ DoCoG: Mask-based Multi-Type Grounded Chain-of-Thought for Document QA https://excv-workshop.github.io/publication/docog-mask-based-multi-type-grounded-chain-of-thought-for-document-qa/ Mon, 01 Jan 0001 00:00:00 +0000 https://excv-workshop.github.io/publication/docog-mask-based-multi-type-grounded-chain-of-thought-for-document-qa/ Emergent Object Binding Has a Finite Spatial Horizon https://excv-workshop.github.io/publication/emergent-object-binding-has-a-finite-spatial-horizon/ Mon, 01 Jan 0001 00:00:00 +0000 https://excv-workshop.github.io/publication/emergent-object-binding-has-a-finite-spatial-horizon/ Evaluating the Interpretability of Sparse Autoencoders with Concept Annotations https://excv-workshop.github.io/publication/evaluating-the-interpretability-of-sparse-autoencoders-with-concept-annotations/ Mon, 01 Jan 0001 00:00:00 +0000 https://excv-workshop.github.io/publication/evaluating-the-interpretability-of-sparse-autoencoders-with-concept-annotations/ EXPL-FR: Explaining Face Recognition Models via Vision-Language Alignment https://excv-workshop.github.io/publication/expl-fr-explaining-face-recognition-models-via-vision-language-alignment/ Mon, 01 Jan 0001 00:00:00 +0000 https://excv-workshop.github.io/publication/expl-fr-explaining-face-recognition-models-via-vision-language-alignment/ EXPOSE: Explainable and Domain-Robust Embeddings from Pathology Vision Foundation Models using Sparse Autoencoders https://excv-workshop.github.io/publication/expose-explainable-and-domain-robust-embeddings-from-pathology-vision-foundation-models-using-sparse-autoencoders/ Mon, 01 Jan 0001 00:00:00 +0000 https://excv-workshop.github.io/publication/expose-explainable-and-domain-robust-embeddings-from-pathology-vision-foundation-models-using-sparse-autoencoders/ Faithful Is Not Trustworthy: The Region of Interest Is Underdetermined in MIL Attribution https://excv-workshop.github.io/publication/faithful-is-not-trustworthy-the-region-of-interest-is-underdetermined-in-mil-attribution/ Mon, 01 Jan 0001 00:00:00 +0000 https://excv-workshop.github.io/publication/faithful-is-not-trustworthy-the-region-of-interest-is-underdetermined-in-mil-attribution/ Formal Concept Lattices are Good Semantic Scaffolds for Concept-Based Learning https://excv-workshop.github.io/publication/formal-concept-lattices-are-good-semantic-scaffolds-for-concept-based-learning/ Mon, 01 Jan 0001 00:00:00 +0000 https://excv-workshop.github.io/publication/formal-concept-lattices-are-good-semantic-scaffolds-for-concept-based-learning/ From Drop-off to Recovery: A Mechanistic Analysis of Segmentation in MLLMs https://excv-workshop.github.io/publication/from-drop-off-to-recovery-a-mechanistic-analysis-of-segmentation-in-mllms/ Mon, 01 Jan 0001 00:00:00 +0000 https://excv-workshop.github.io/publication/from-drop-off-to-recovery-a-mechanistic-analysis-of-segmentation-in-mllms/ From Image Latent Space to Fuzzy Rules: Interpretable Analysis of Gastrointestinal Foundation Model https://excv-workshop.github.io/publication/from-image-latent-space-to-fuzzy-rules-interpretable-analysis-of-gastrointestinal-foundation-model/ Mon, 01 Jan 0001 00:00:00 +0000 https://excv-workshop.github.io/publication/from-image-latent-space-to-fuzzy-rules-interpretable-analysis-of-gastrointestinal-foundation-model/ From Interpretability Methods to Interpretable Models https://excv-workshop.github.io/publication/from-interpretability-methods-to-interpretable-models/ Mon, 01 Jan 0001 00:00:00 +0000 https://excv-workshop.github.io/publication/from-interpretability-methods-to-interpretable-models/ Influence Functions for Flow-Matching Generative Models: A K-FAC Port and an Honest Scale Study https://excv-workshop.github.io/publication/influence-functions-for-flow-matching-generative-models-a-k-fac-port-and-an-honest-scale-study/ Mon, 01 Jan 0001 00:00:00 +0000 https://excv-workshop.github.io/publication/influence-functions-for-flow-matching-generative-models-a-k-fac-port-and-an-honest-scale-study/ Learn to Rank: Visual Attribution by Learning Importance Ranking https://excv-workshop.github.io/publication/learn-to-rank-visual-attribution-by-learning-importance-ranking/ Mon, 01 Jan 0001 00:00:00 +0000 https://excv-workshop.github.io/publication/learn-to-rank-visual-attribution-by-learning-importance-ranking/ LogicCBMs: Logic-Enhanced Concept-Based Learning https://excv-workshop.github.io/publication/logiccbms-logic-enhanced-concept-based-learning/ Mon, 01 Jan 0001 00:00:00 +0000 https://excv-workshop.github.io/publication/logiccbms-logic-enhanced-concept-based-learning/ Logit Lens Supervision for Patch-Level Explanations in Vision-Language Models https://excv-workshop.github.io/publication/logit-lens-supervision-for-patch-level-explanations-in-vision-language-models/ Mon, 01 Jan 0001 00:00:00 +0000 https://excv-workshop.github.io/publication/logit-lens-supervision-for-patch-level-explanations-in-vision-language-models/ Modeling Visual Flow Leakage to Alleviate Hallucination in Large Vision-Language Models https://excv-workshop.github.io/publication/modeling-visual-flow-leakage-to-alleviate-hallucination-in-large-vision-language-models/ Mon, 01 Jan 0001 00:00:00 +0000 https://excv-workshop.github.io/publication/modeling-visual-flow-leakage-to-alleviate-hallucination-in-large-vision-language-models/ On the Faithfulness of Post-Hoc Concept Bottleneck Models https://excv-workshop.github.io/publication/on-the-faithfulness-of-post-hoc-concept-bottleneck-models/ Mon, 01 Jan 0001 00:00:00 +0000 https://excv-workshop.github.io/publication/on-the-faithfulness-of-post-hoc-concept-bottleneck-models/ Position: We need theory-grounded explainability for abstract tasks to quantify model bias. https://excv-workshop.github.io/publication/position-we-need-theory-grounded-explainability-for-abstract-tasks-to-quantify-model-bias./ Mon, 01 Jan 0001 00:00:00 +0000 https://excv-workshop.github.io/publication/position-we-need-theory-grounded-explainability-for-abstract-tasks-to-quantify-model-bias./ ProtoMappingNet: Interpretable Hierarchical Prototypes through Relational Prototype Mappings https://excv-workshop.github.io/publication/protomappingnet-interpretable-hierarchical-prototypes-through-relational-prototype-mappings/ Mon, 01 Jan 0001 00:00:00 +0000 https://excv-workshop.github.io/publication/protomappingnet-interpretable-hierarchical-prototypes-through-relational-prototype-mappings/ Queryable Is Not Grounded: Reference-Anchored Faithfulness Evaluation for Foundation Vision-Language Models https://excv-workshop.github.io/publication/queryable-is-not-grounded-reference-anchored-faithfulness-evaluation-for-foundation-vision-language-models/ Mon, 01 Jan 0001 00:00:00 +0000 https://excv-workshop.github.io/publication/queryable-is-not-grounded-reference-anchored-faithfulness-evaluation-for-foundation-vision-language-models/ Seeing Before Answering: Training-Free Visual Layer Profiling for Vision-Language Models https://excv-workshop.github.io/publication/seeing-before-answering-training-free-visual-layer-profiling-for-vision-language-models/ Mon, 01 Jan 0001 00:00:00 +0000 https://excv-workshop.github.io/publication/seeing-before-answering-training-free-visual-layer-profiling-for-vision-language-models/ SIC: Similarity-Based Interpretable Image Classification with Neural Networks https://excv-workshop.github.io/publication/sic-similarity-based-interpretable-image-classification-with-neural-networks/ Mon, 01 Jan 0001 00:00:00 +0000 https://excv-workshop.github.io/publication/sic-similarity-based-interpretable-image-classification-with-neural-networks/ Vision-Language Models Trust What They Can Discern: Image and Representational Properties Governing Modality Arbitration https://excv-workshop.github.io/publication/vision-language-models-trust-what-they-can-discern-image-and-representational-properties-governing-modality-arbitration/ Mon, 01 Jan 0001 00:00:00 +0000 https://excv-workshop.github.io/publication/vision-language-models-trust-what-they-can-discern-image-and-representational-properties-governing-modality-arbitration/ VisLens: Single-Pass Interpretable Visual Search for Multimodal LLMs https://excv-workshop.github.io/publication/vislens-single-pass-interpretable-visual-search-for-multimodal-llms/ Mon, 01 Jan 0001 00:00:00 +0000 https://excv-workshop.github.io/publication/vislens-single-pass-interpretable-visual-search-for-multimodal-llms/ Why Far Looks Up: Probing Spatial Representation in Vision-Language Models https://excv-workshop.github.io/publication/why-far-looks-up-probing-spatial-representation-in-vision-language-models/ Mon, 01 Jan 0001 00:00:00 +0000 https://excv-workshop.github.io/publication/why-far-looks-up-probing-spatial-representation-in-vision-language-models/