Assessing Multimodal Chronic Wound Embeddings with Expert Triplet Agreement

Recessive dystrophic epidermolysis bullosa (RDEB) is a rare genetic skin disorder for which clinicians greatly benefit from finding similar cases using images and clinical text. However, off-the-shelf foundation models do not reliably capture clinically meaningful features for this heterogeneous, long-tail disease, and structured measurement of agreement with experts is challenging. To address these gaps, we propose evaluating embedding spaces with expert ordinal comparisons (triplet judgments), which are fast to collect and encode implicit clinical similarity knowledge. We further introduce TriDerm, a multimodal framework that learns interpretable wound representations from small cohorts by integrating wound imagery, boundary masks, and expert reports. On the vision side, TriDerm adapts visual foundation models to RDEB using wound-level attention pooling and non-contrastive representation learning. For text, we prompt large language models with comparison queries and recover medically meaningful representations via soft ordinal embeddings (SOE). We show that visual and textual modalities capture complementary aspects of wound phenotype, and that fusing both modalities yields 73.5% agreement with experts, outperforming the best off-the-shelf single-modality foundation model by over 5.6 percentage points. We make the expert annotation tool, model code and representative dataset samples publicly available.

Embedding interpretable $\ell_1$-regression into neural networks for uncovering temporal structure in cell imaging

While artificial neural networks excel in unsupervised learning of non-sparse structure, classical statistical regression techniques offer better interpretability, in particular when sparseness is enforced by $\ell_1$ regularization, enabling identification of which factors drive observed dynamics. We investigate how these two types of approaches can be optimally combined, exemplarily considering two-photon calcium imaging data where sparse autoregressive dynamics are to be extracted. We propose embedding a vector autoregressive (VAR) model as an interpretable regression technique into a convolutional autoencoder, which provides dimension reduction for tractable temporal modeling. A skip connection separately addresses non-sparse static spatial information, selectively channeling sparse structure into the $\ell_1$-regularized VAR. $\ell_1$-estimation of regression parameters is enabled by differentiating through the piecewise linear solution path. This is contrasted with approaches where the autoencoder does not adapt to the VAR model. Having an embedded statistical model also enables a testing approach for comparing temporal sequences from the same observational unit. Additionally, contribution maps visualize which spatial regions drive the learned dynamics.

Learning to Read Where to Look: Disease-Aware Vision-Language Pretraining for 3D CT

Recent 3D CT vision-language models align volumes with reports via contrastive pretraining, but typically rely on limited public data and provide only coarse global supervision. We train a 3D CT vision-language model on 98k report-volume pairs (50k patients) collected at a single hospital, combined with public datasets, using SigLIP-style contrastive pretraining together with prompt-based disease supervision in the shared vision-text embedding space. On CT-RATE, our model achieves state-of-the-art text-to-image retrieval (R@10 31.5 vs. 22.2) and competitive disease classification (AUC 83.8 vs. 83.8), with consistent results on Rad-ChestCT (AUC 77.0 vs. 77.3). We further observe that radiologists routinely reference specific images within their reports (e.g., ``series X, image Y''), linking textual descriptions to precise axial locations. We automatically mine 262k such snippet-slice pairs and introduce the task of intra-scan snippet localization -- predicting the axial depth referred to by a text snippet -- reducing mean absolute error to 36.3 mm at 12 mm feature resolution, compared with 67.0 mm for the best baseline. Adding this localization objective leaves retrieval and classification broadly unchanged within confidence bounds, yielding a single unified model for retrieval, classification, and intra-scan grounding.

Simple LLM Baselines are Competitive for Model Diffing

Standard LLM evaluations only test capabilities or dispositions that evaluators designed them for, missing unexpected differences such as behavioral shifts between model revisions or emergent misaligned tendencies. Model diffing addresses this limitation by automatically surfacing systematic behavioral differences. Recent approaches include LLM-based methods that generate natural language descriptions and sparse autoencoder (SAE)-based methods that identify interpretable features. However, no systematic comparison of these approaches exists nor are there established evaluation criteria. We address this gap by proposing evaluation metrics for key desiderata (generalization, interestingness, and abstraction level) and use these to compare existing methods. Our results show that an improved LLM-based baseline performs comparably to the SAE-based method while typically surfacing more abstract behavioral differences.

On Geometric Understanding and Learned Data Priors in VGGT

The Visual Geometry Grounded Transformer (VGGT) is a 3D foundation model that infers camera geometry and scene structure in a single feed-forward pass. Trained in a supervised, single-step fashion on large datasets, VGGT raises a key question: does it build upon geometric concepts like traditional multi-view methods, or does it rely primarily on learned appearance-based data-driven priors? In this work, we conduct a systematic analysis of VGGT's internal mechanisms to uncover whether geometric understanding emerges within its representations. By probing intermediate features, analyzing attention patterns, and performing interventions, we examine how the model implements its functionality. Our findings reveal that VGGT implicitly performs correspondence matching within its global attention layers and encodes epipolar geometry, despite being trained without explicit geometric constraints. We further investigate VGGT's dependence on its learned data priors. Using spatial input masking and perturbation experiments, we assess its robustness to occlusions, appearance variations, and camera configurations, comparing it with classical multi-stage pipelines. Together, these insights highlight how VGGT internalizes geometric structure while using learned data-driven priors.

Orbis: Overcoming Challenges of Long-Horizon Prediction in Driving World Models

Existing world models for autonomous driving struggle with long-horizon generation and generalization to challenging scenarios. In this work, we develop a model using simple design choices, and without additional supervision or sensors, such as maps, depth, or multiple cameras. We show that our model yields state-of-the-art performance, despite having only 469M parameters and being trained on 280h of video data. It particularly stands out in difficult scenarios like turning maneuvers and urban traffic. We test whether discrete token models possibly have advantages over continuous models based on flow matching. To this end, we set up a hybrid tokenizer that is compatible with both approaches and allows for a side-by-side comparison. Our study concludes in favor of the continuous autoregressive model, which is less brittle on individual design choices and more powerful than the model built on discrete tokens. Code, models and qualitative results are publicly available at https://lmb-freiburg.github.io/orbis.github.io/.

CLIP Won't Learn Object-Attribute Binding from Natural Data and Here is Why

Contrastive vision-language models like CLIP are used for a large variety of applications, such as zero-shot classification or as vision encoder for multi-modal models. Despite their popularity, their representations show major limitations. For instance, CLIP models learn bag-of-words representations and, as a consequence, fail to distinguish whether an image is of "a yellow submarine and a blue bus" or "a blue submarine and a yellow bus". Previous attempts to fix this issue added hard negatives during training or modified the architecture, but failed to resolve the problem in its entirety. We suspect that the missing insights to solve the binding problem for CLIP are hidden in the arguably most important part of learning algorithms: the data. In this work, we fill this gap by rigorously identifying the influence of data properties on CLIP's ability to learn binding using a synthetic dataset. We find that common properties of natural data such as low attribute density, incomplete captions, and the saliency bias, a tendency of human captioners to describe the object that is "most salient" to them have a detrimental effect on binding performance. In contrast to common belief, we find that neither scaling the batch size, i.e., implicitly adding more hard negatives, nor explicitly creating hard negatives enables CLIP to learn reliable binding. Only when the data expresses our identified data properties CLIP learns almost perfect binding.

cVLA: Towards Efficient Camera-Space VLAs

Vision-Language-Action (VLA) models offer a compelling framework for tackling complex robotic manipulation tasks, but they are often expensive to train. In this paper, we propose a novel VLA approach that leverages the competitive performance of Vision Language Models (VLMs) on 2D images to directly infer robot end-effector poses in image frame coordinates. Unlike prior VLA models that output low-level controls, our model predicts trajectory waypoints, making it both more efficient to train and robot embodiment agnostic. Despite its lightweight design, our next-token prediction architecture effectively learns meaningful and executable robot trajectories. We further explore the underutilized potential of incorporating depth images, inference-time techniques such as decoding strategies, and demonstration-conditioned action generation. Our model is trained on a simulated dataset and exhibits strong sim-to-real transfer capabilities. We evaluate our approach using a combination of simulated and real data, demonstrating its effectiveness on a real robotic system.

Using Knowledge Graphs to harvest datasets for efficient CLIP model training

Training high-quality CLIP models typically requires enormous datasets, which limits the development of domain-specific models -- especially in areas that even the largest CLIP models do not cover well -- and drives up training costs. This poses challenges for scientific research that needs fine-grained control over the training procedure of CLIP models. In this work, we show that by employing smart web search strategies enhanced with knowledge graphs, a robust CLIP model can be trained from scratch with considerably less data. Specifically, we demonstrate that an expert foundation model for living organisms can be built using just 10M images. Moreover, we introduce EntityNet, a dataset comprising 33M images paired with 46M text descriptions, which enables the training of a generic CLIP model in significantly reduced time.

Detect, Classify, Act: Categorizing Industrial Anomalies with Multi-Modal Large Language Models

Recent advances in visual industrial anomaly detection have demonstrated exceptional performance in identifying and segmenting anomalous regions while maintaining fast inference speeds. However, anomaly classification-distinguishing different types of anomalies-remains largely unexplored despite its critical importance in real-world inspection tasks. To address this gap, we propose VELM, a novel LLM-based pipeline for anomaly classification. Given the critical importance of inference speed, we first apply an unsupervised anomaly detection method as a vision expert to assess the normality of an observation. If an anomaly is detected, the LLM then classifies its type. A key challenge in developing and evaluating anomaly classification models is the lack of precise annotations of anomaly classes in existing datasets. To address this limitation, we introduce MVTec-AC and VisA-AC, refined versions of the widely used MVTec-AD and VisA datasets, which include accurate anomaly class labels for rigorous evaluation. Our approach achieves a state-of-the-art anomaly classification accuracy of 80.4% on MVTec-AD, exceeding the prior baselines by 5%, and 84% on MVTec-AC, demonstrating the effectiveness of VELM in understanding and categorizing anomalies. We hope our methodology and benchmark inspire further research in anomaly classification, helping bridge the gap between detection and comprehensive anomaly characterization.