@article{20934, abstract = { Supervised learning for causal discovery from observational data often achieves competitive performance despite seemingly avoiding the explicit assumptions that traditional methods require for identifiability. In this work, we analyze CSIvA (Ke et al., 2023) on bivariate causal models, a transformer architecture for amortized inference promising to train on synthetic data and transfer to real ones. First, we bridge the gap with identifiability theory, showing that the training distribution implicitly defines a prior on the causal model of the test observations: consistent with classical approaches, good performance is achieved when we have a good prior on the test data, and the underlying model is identifiable. Second, we find that CSIvA can not generalize to classes of causal models unseen during training: to overcome this limitation, we theoretically and empirically analyze \textit{when} training CSIvA on datasets generated by multiple identifiable causal models with different structural assumptions improves its generalization at test time. Overall, we find that amortized causal discovery still adheres to identifiability theory, violating the previous hypothesis from Lopez-Paz et al. (2015) that supervised learning methods could overcome its restrictions.}, author = {Montagna, Francesco and Cairney-Leeming, Maximilian T and Sridhar, Dhanya and Locatello, Francesco}, issn = {2835-8856}, journal = {Transactions on Machine Learning Research}, publisher = {ML Research Press}, title = {{Demystifying amortized causal discovery with transformers}}, year = {2025}, } @inproceedings{21049, abstract = {Post-hoc importance attribution methods are a popular tool for “explaining” Deep Neural Networks (DNNs) and are inherently based on the assumption that the explanations can be applied independently of how the models were trained. Contrarily, in this work we bring forward empirical evidence that challenges this very notion. Surprisingly, we discover a strong dependency on and demonstrate that the training details of a pre-trained model’s classification layer (<10% of model parameters) play a crucial role, much more than the pre-training scheme itself. This is of high practical relevance: (1) as techniques for pre-training models are becoming increasingly diverse, understanding the interplay between these techniques and attribution methods is critical; (2) it sheds light on an important yet overlooked assumption of post-hoc attribution methods which can drastically impact model explanations and how they are interpreted eventually. With this finding we also present simple yet effective adjustments to the classification layers, that can significantly enhance the quality of model explanations. We validate our findings across several visual pre-training frameworks (fully-supervised, self-supervised, contrastive vision-language training) and analyse how they impact explanations for a wide range of attribution methods on a diverse set of evaluation metrics.}, author = {Gairola, Siddhartha and Böhle, Moritz and Locatello, Francesco and Schiele, Bernt}, booktitle = {13th International Conference on Learning Representations}, location = {Singapore}, publisher = {ICLR}, title = {{How to probe: Simple yet effective techniques for improving post-hoc explanations}}, year = {2025}, } @inproceedings{21066, abstract = {Causal discovery from observational data holds great promise, but existing methods rely on strong assumptions about the underlying causal structure, often requiring full observability of all relevant variables. We tackle these challenges by leveraging the score function ∇logp(X) of observed variables for causal discovery and propose the following contributions. First, we generalize the existing results of identifiability with the score to additive noise models with minimal requirements on the causal mechanisms. Second, we establish conditions for inferring causal relations from the score even in the presence of hidden variables; this result is two-faced: we demonstrate the score’s potential as an alternative to conditional independence tests to infer the equivalence class of causal graphs with hidden variables, and we provide the necessary conditions for identifying direct causes in latent variable models. Building on these insights, we propose a flexible algorithm for causal discovery across linear, nonlinear, and latent variable models, which we empirically validate.}, author = {Montagna, Francesco and Faller, Philipp and Blöbaum, Patrik and Kirschbaum, Elke and Locatello, Francesco}, booktitle = {Proceedings of the Fourth Conference on Causal Learning and Reasoning}, issn = {2640-3498}, location = {Lausanne, Switzerland}, pages = {552--605}, publisher = {ML Research Press}, title = {{Score matching through the roof: Linear, nonlinear, and latent variables causal discovery}}, volume = {275}, year = {2025}, } @inproceedings{20032, abstract = {We propose Scalable Mechanistic Neural Network (S-MNN), an enhanced neural network framework designed for scientific machine learning applications involving long temporal sequences. By reformulating the original Mechanistic Neural Network (MNN) (Pervez et al., 2024), we reduce the computational time and space complexities from cubic and quadratic with respect to the sequence length, respectively, to linear. This significant improvement enables efficient modeling of long-term dynamics without sacrificing accuracy or interpretability. Extensive experiments demonstrate that S-MNN matches the original MNN in precision while substantially reducing computational resources. Consequently, S-MNN can drop-in replace the original MNN in applications, providing a practical and efficient tool for integrating mechanistic bottlenecks into neural network models of complex dynamical systems. Source code is available at https://github.com/IST-DASLab/ScalableMNN.}, author = {Chen, Jiale and Yao, Dingling and Pervez, Adeel A and Alistarh, Dan-Adrian and Locatello, Francesco}, booktitle = {13th International Conference on Learning Representations}, isbn = {9798331320850}, location = {Singapore, Singapore}, pages = {63716--63737}, publisher = {ICLR}, title = {{Scalable mechanistic neural networks}}, year = {2025}, } @inproceedings{20036, abstract = {We introduce NeCo: Patch Neighbor Consistency, a novel self-supervised training loss that enforces patch-level nearest neighbor consistency across a student and teacher model. Compared to contrastive approaches that only yield binary learning signals, i.e. "attract" and "repel", this approach benefits from the more fine-grained learning signal of sorting spatially dense features relative to reference patches. Our method leverages differentiable sorting applied on top of pretrained representations, such as DINOv2-registers to bootstrap the learning signal and further improve upon them. This dense post-pretraining leads to superior performance across various models and datasets, despite requiring only 19 hours on a single GPU. This method generates high-quality dense feature encoders and establishes several new state-of-the-art results such as +2.3 % and +4.2% for non-parametric in-context semantic segmentation on ADE20k and Pascal VOC, +1.6% and +4.8% for linear segmentation evaluations on COCO-Things and -Stuff and improvements in the 3D understanding of multi-view consistency on SPair-71k, by more than 1.5%.}, author = {Pariza, Valentinos and Salehi, Mohammadreza and Burghouts, Gertjan and Locatello, Francesco and Asano, Yuki M.}, booktitle = {13th International Conference on Learning Representations}, isbn = {9798331320850}, location = {Singapore, Singapore}, pages = {72303--72330}, publisher = {ICLR}, title = {{Near, far: Patch-ordering enhances vision foundation models' scene understanding}}, year = {2025}, } @unpublished{19674, abstract = {When examined through the lens of their residual streams, a puzzling property emerges in transformer networks: residual contributions (e.g., attention heads) sometimes specialize in specific tasks or input attributes. In this paper, we analyze this phenomenon in vision transformers, focusing on the spectral geometry of residuals, and explore its implications for modality alignment in vision-language models. First, we link it to the intrinsically low-dimensional structure of visual head representations, zooming into their principal components and showing that they encode specialized roles across a wide variety of input data distributions. Then, we analyze the effect of head specialization in multimodal models, focusing on how improved alignment between text and specialized heads impacts zero-shot classification performance. This specialization-performance link consistently holds across diverse pre-training data, network sizes, and objectives, demonstrating a powerful new mechanism for boosting zero-shot classification through targeted alignment. Ultimately, we translate these insights into actionable terms by introducing ResiDual, a technique for spectral alignment of the residual stream. Much like panning for gold, it lets the noise from irrelevant unit principal components (i.e., attributes) wash away to amplify task-relevant ones. Remarkably, this dual perspective on modality alignment yields fine-tuning level performance on different data distributions while modelling an extremely interpretable and parameter-efficient transformation, as we extensively show on 70 pre-trained network-dataset combinations (7 models, 10 datasets).}, author = {Basile, Lorenzo and Maiorca, Valentino and Bortolussi, Luca and Rodolà, Emanuele and Locatello, Francesco}, booktitle = {arXiv}, title = {{ResiDual transformer alignment with spectral decomposition}}, doi = {10.48550/arXiv.2411.00246}, year = {2025}, } @inproceedings{20817, abstract = {We present Mechanistic PDE Networks -- a model for discovery of governing partial differential equations from data. Mechanistic PDE Networks represent spatiotemporal data as space-time dependent linear partial differential equations in neural network hidden representations. The represented PDEs are then solved and decoded for specific tasks. The learned PDE representations naturally express the spatiotemporal dynamics in data in neural network hidden space, enabling increased modeling power. Solving the PDE representations in a compute and memory-efficient way, however, is a significant challenge. We develop a native, GPU-capable, parallel, sparse and differentiable multigrid solver specialized for linear partial differential equations that acts as a module in Mechanistic PDE Networks. Leveraging the PDE solver we propose a discovery architecture that can discovers nonlinear PDEs in complex settings, while being robust to noise. We validate PDE discovery on a number of PDEs including reaction-diffusion and Navier-Stokes equations.}, author = {Pervez, Adeel A and Gavves, Efstratios and Locatello, Francesco}, booktitle = {42nd International Conference on Machine Learning}, issn = {2640-3498}, location = {Vancouver, Canada}, pages = {48962--48973}, publisher = {ML Research Press}, title = {{Mechanistic PDE networks for discovery of governing equations}}, volume = {267}, year = {2025}, } @inproceedings{22832, abstract = {In many scientific experiments, the data annotating cost constraints the pace for testing novel hypotheses. Yet, modern machine learning pipelines offer a promising solution—provided their predictions yield correct conclusions. We focus on Prediction-Powered Causal Inferences (PPCI), i.e., estimating the treatment effect in an unlabeled target experiment, relying on training data with the same outcome annotated but potentially different treatment or effect modifiers. We first show that conditional calibration guarantees valid PPCI at population level. Then, we introduce a sufficient representation constraint transferring validity across experiments, which we propose to enforce in practice in Deconfounded Empirical Risk Minimization, our new model-agnostic training objective. We validate our method on synthetic and real-world scientific data, solving impossible problem instances for Empirical Risk Minimization even with standard invariance constraints. In particular, for the first time, we achieve valid causal inference on a scientific experiment with complex recording and no human annotations, fine-tuning a foundational model on our similar annotated experiment.}, author = {Cadei, Riccardo and Demirel, Ilker and De Bartolomeis, Piersilvio and Lindorfer, Lukas and Cremer, Sylvia and Schmid, Cordelia and Locatello, Francesco}, booktitle = {39th Conference on Neural Information Processing Systems}, isbn = {9798331338275}, issn = {1049-5258}, location = {San Diego, CA, United States}, pages = {82200--82229}, publisher = {Neural Information Processing Systems Foundation}, title = {{Prediction-Powered Causal Inferences}}, doi = {10.52202/085713-2479}, volume = {38}, year = {2025}, } @inproceedings{22831, abstract = {Neural models learn representations of high-dimensional data on low-dimensional manifolds. Multiple factors, including stochasticities in the training process, model architectures, and additional inductive biases, may induce different representations, even when learning the same task on the same data. However, it has recently been shown that when a latent structure is shared between distinct latent spaces, relative distances between representations can be preserved, up to distortions. Building on this idea, we demonstrate that exploiting the differential-geometric structure of latent spaces of neural models, it is possible to capture precisely the transformations between representational spaces trained on similar data distributions. Specifically, we assume that distinct neural models parametrize approximately the same underlying manifold, and introduce a representation based on the pullback metric that captures the intrinsic structure of the latent space, while scaling efficiently to large models. We validate experimentally our method on model stitching and retrieval tasks, covering autoencoders and vision foundation discriminative models, across diverse architectures, datasets, pretraining schemes and modalities. Code is available at https://github.com/marc0git/RelativeGeodesics.}, author = {Yu, Hanlin and Inal, Berfin and Arvanitidis, Georgios and Hauberg, Søren and Locatello, Francesco and Fumero, Marco}, booktitle = {39th Conference on Neural Information Processing Systems}, isbn = {9798331338275}, issn = {1049-5258}, location = {San Diego, CA, United States}, pages = {125316--125360}, publisher = {Neural Information Processing Systems Foundation}, title = {{Connecting neural models latent geometries with relative geodesic representations}}, doi = {10.52202/085713-3769}, volume = {38}, year = {2025}, } @inproceedings{19010, abstract = {Causal representation learning aims at recovering latent causal variables from high-dimensional observations to solve causal downstream tasks, such as predicting the effect of new interventions or more robust classification. A plethora of methods have been developed, each tackling carefully crafted problem settings that lead to different types of identifiability. The folklore is that these different settings are important, as they are often linked to different rungs of Pearl's causal hierarchy, although not all neatly fit. Our main contribution is to show that many existing causal representation learning approaches methodologically align the representation to known data symmetries. Identification of the variables is guided by equivalence classes across different "data pockets" that are not necessarily causal. This result suggests important implications, allowing us to unify many existing approaches in a single method that can mix and match different assumptions, including non-causal ones, based on the invariances relevant to our application. It also significantly benefits applicability, which we demonstrate by improving treatment effect estimation on real-world high-dimensional ecological data. Overall, this paper clarifies the role of causality assumptions in the discovery of causal variables and shifts the focus to preserving data symmetries.}, author = {Yao, Dingling and Rancati, Dario and Cadei, Riccardo and Fumero, Marco and Locatello, Francesco}, booktitle = {13th International Conference on Learning Representations}, location = {Singapore}, publisher = {ICLR}, title = {{Unifying causal representation learning with the invariance principle}}, year = {2025}, } @inproceedings{21074, abstract = {Neural models learn representations of high-dimensional data on low-dimensional manifolds. Multiple factors, including stochasticities in the training process, model architectures, and additional inductive biases, may induce different representations, even when learning the same task on the same data. However, it has recently been shown that when a latent structure is shared between distinct latent spaces, relative distances between representations can be preserved, up to distortions. Building on this idea, we demonstrate that exploiting the differential-geometric structure of latent spaces of neural models, it is possible to capture precisely the transformations between representational spaces trained on similar data distributions. Specifically, we assume that distinct neural models parametrize approximately the same underlying manifold, and introduce a representation based on the pullback metric that captures the intrinsic structure of the latent space, while scaling efficiently to large models. We validate experimentally our method on model stitching and retrieval tasks, covering autoencoders and vision foundation discriminative models, across diverse architectures, datasets, pretraining schemes and modalities. Code is available at the following link.}, author = {Yu, Hanlin and Inal, Befrin and Arvanitidis, Georgios and Hauberg, Soren and Locatello, Francesco and Fumero, Marco}, booktitle = {39th Annual Conference on Neural Information Processing Systems}, issn = {1049-5258}, location = {San Diego, CA, United States}, publisher = {Neural Information Processing Systems Foundation}, title = {{Connecting neural models latent geometries with relative geodesic representations}}, doi = {10.52202/085713-3769}, volume = {38}, year = {2025}, } @inproceedings{21076, abstract = {In many scientific experiments, the data annotating cost constraints the pace for testing novel hypotheses. Yet, modern machine learning pipelines offer a promising solution—provided their predictions yield correct conclusions. We focus on Prediction-Powered Causal Inferences (PPCI), i.e., estimating the treatment effect in an unlabeled target experiment, relying on training data with the same outcome annotated but potentially different treatment or effect modifiers. We first show that conditional calibration guarantees valid PPCI at population level. Then, we introduce a sufficient representation constraint transferring validity across experiments, which we propose to enforce in practice in Deconfounded Empirical Risk Minimization, our new model-agnostic training objective. We validate our method on synthetic and real-world scientific data, solving impossible problem instances for Empirical Risk Minimization even with standard invariance constraints. In particular, for the first time, we achieve valid causal inference on a scientific experiment with complex recording and no human annotations, fine-tuning a foundational model on our similar annotated experiment.}, author = {Cadei, Riccardo and Demirel, Ilker and De Bartolomeis, Piersilvio and Lindorfer, Lukas and Cremer, Sylvia and Schmid, Cordelia and Locatello, Francesco}, booktitle = {39th Annual Conference on Neural Information Processing Systems}, issn = {1049-5258}, location = {San Diego, CA, United States}, publisher = {Neural Information Processing Systems Foundation}, title = {{Prediction-powered causal inferences}}, doi = {10.52202/085713-2479}, volume = {38}, year = {2025}, } @inproceedings{21072, abstract = {Language and vision-language models have shown impressive performance across a wide range of tasks, but their internal mechanisms remain only partly understood. In this work, we study how individual attention heads in text-generative models specialize in specific semantic or visual attributes. Building on an established interpretability method, we reinterpret the practice of probing intermediate activations with the final decoding layer through the lens of signal processing. This lets us analyze multiple samples in a principled way and rank attention heads based on their relevance to target concepts. Our results show consistent patterns of specialization at the head level across both unimodal and multimodal transformers. Remarkably, we find that editing as few as 1% of the heads, selected using our method, can reliably suppress or enhance targeted concepts in the model output. We validate our approach on language tasks such as question answering and toxicity mitigation, as well as vision-language tasks including image classification and captioning. Our findings highlight an interpretable and controllable structure within attention layers, offering simple tools for understanding and editing large-scale generative models.}, author = {Basile, Lorenzo and Maiorca, Valentino and Doimo, Diego and Locatello, Francesco and Cazzaniga, Alberto}, booktitle = {39th Annual Conference on Neural Information Processing Systems}, issn = {1049-5258}, location = {San Diego, CA, United States}, publisher = {Neural Information Processing Systems Foundation}, title = {{Head pursuit: Probing attention specialization in multimodal transformers}}, doi = {10.52202/085713-3752}, volume = {38}, year = {2025}, } @inproceedings{21068, abstract = {Causal reasoning and discovery, two fundamental tasks of causal analysis, often face challenges in applications due to the complexity, noisiness, and highdimensionality of real-world data. Despite recent progress in identifying latent causal structures using causal representation learning (CRL), what makes learned representations useful for causal downstream tasks and how to evaluate them are still not well understood. In this paper, we reinterpret CRL using a measurement model framework, where the learned representations are viewed as proxy measurements of the latent causal variables. Our approach clarifies the conditions under which learned representations support downstream causal reasoning and provides a principled basis for quantitatively assessing the quality of representations using a new Test-based Measurement EXclusivity (T-MEX) score. We validate T-MEX across diverse causal inference scenarios, including numerical simulations and real-world ecological video analysis, demonstrating that the proposed framework and corresponding score effectively assess the identification of learned representations and their usefulness for causal downstream tasks. Reproducible code can be found at https://github.com/shimenghuang/a-measurement-perspective-of-crl.}, author = {Yao, Dingling and Huang, Shimeng and Cadei, Riccardo and Zhang, Kun and Locatello, Francesco}, booktitle = {39th Annual Conference on Neural Information Processing Systems}, issn = {1049-5258}, location = {San Diego, CA, United States}, publisher = {Neural Information Processing Systems Foundation}, title = {{The third pillar of causal analysis? A measurement perspective on causal representations}}, doi = {10.52202/085713-1444}, volume = {38}, year = {2025}, } @inproceedings{21070, abstract = {Deep learning systems deployed in real-world applications often encounter data that is different from their in-distribution (ID). A reliable model should ideally abstain from making decisions in this out-of-distribution (OOD) setting. Existing state-of-the-art methods primarily focus on feature distances, such as k-th nearest neighbors and distances to decision boundaries, either overlooking or ineffectively using in-distribution statistics. In this work, we propose a novel angle-based metric for OOD detection that is computed relative to the in-distribution structure. We demonstrate that the angles between feature representations and decision boundaries, viewed from the mean of in-distribution features, serve as an effective discriminative factor between ID and OOD data. We evaluate our method on nine ImageNet-pretrained models. Our approach achieves the lowest FPR in 5 out of 9 ImageNet models, obtains the best average FPR overall, and consistently ranking among the top 3 across all evaluated models. Furthermore, we highlight the benefits of contrastive representations by showing strong performance with ResNet SCL and CLIP architectures. Finally, we demonstrate that the scale-invariant nature of our score enables an ensemble strategy via simple score summation. }, author = {Demirel, Berker and Fumero, Marco and Locatello, Francesco}, booktitle = {39th Annual Conference on Neural Information Processing Systems}, issn = {1049-5258}, location = {San Diego, CA, United States}, publisher = {Neural Information Processing Systems Foundation}, title = {{Out-of-Distribution detection with relative angles}}, doi = {10.52202/085713-3125}, volume = {38}, year = {2025}, } @inproceedings{14213, abstract = {We introduce a method to segment the visual field into independently moving regions, trained with no ground truth or supervision. It consists of an adversarial conditional encoder-decoder architecture based on Slot Attention, modified to use the image as context to decode optical flow without attempting to reconstruct the image itself. In the resulting multi-modal representation, one modality (flow) feeds the encoder to produce separate latent codes (slots), whereas the other modality (image) conditions the decoder to generate the first (flow) from the slots. This design frees the representation from having to encode complex nuisance variability in the image due to, for instance, illumination and reflectance properties of the scene. Since customary autoencoding based on minimizing the reconstruction error does not preclude the entire flow from being encoded into a single slot, we modify the loss to an adversarial criterion based on Contextual Information Separation. The resulting min-max optimization fosters the separation of objects and their assignment to different attention slots, leading to Divided Attention, or DivA. DivA outperforms recent unsupervised multi-object motion segmentation methods while tripling run-time speed up to 104FPS and reducing the performance gap from supervised methods to 12% or less. DivA can handle different numbers of objects and different image sizes at training and test time, is invariant to permutation of object labels, and does not require explicit regularization.}, author = {Lao, Dong and Hu, Zhengyang and Locatello, Francesco and Yang, Yanchao and Soatto, Stefano}, booktitle = {1st Conference on Parsimony and Learning}, location = {Hong Kong, China}, title = {{Divided attention: Unsupervised multi-object discovery with contextually separated slots}}, year = {2024}, } @inproceedings{18114, abstract = {This paper presents Mechanistic Neural Networks, a neural network design for machine learning applications in the sciences. It incorporates a new Mechanistic Block in standard architectures to explicitly learn governing differential equations as representations, revealing the underlying dynamics of data and enhancing interpretability and efficiency in data modeling. Central to our approach is a novel Relaxed Linear Programming Solver (NeuRLP) inspired by a technique that reduces solving linear ODEs to solving linear programs. This integrates well with neural networks and surpasses the limitations of traditional ODE solvers enabling scalable GPU parallel processing. Overall, Mechanistic Neural Networks demonstrate their versatility for scientific machine learning applications, adeptly managing tasks from equation discovery to dynamic systems modeling. We prove their comprehensive capabilities in analyzing and interpreting complex scientific data across various applications, showing significant performance against specialized state-of-the-art methods. Source code is available at https://github.com/alpz/mech-nn.}, author = {Pervez, Adeel A and Locatello, Francesco and Gavves, Efstratios}, booktitle = {Proceedings of the 41st International Conference on Machine Learning}, issn = {2640-3498}, location = {Vienna, Austria}, pages = {40484--40501}, publisher = {ML Research Press}, title = {{Mechanistic neural networks for scientific machine learning}}, volume = {235}, year = {2024}, } @inproceedings{18964, abstract = {Object-centric learning (OCL) extracts the representation of objects with slots, offering an exceptional blend of flexibility and interpretability for abstracting low-level perceptual features. A widely adopted method within OCL is slot attention, which utilizes attention mechanisms to iteratively refine slot representations. However, a major draw-back of most object-centric models, including slot attention, is their reliance on predefining the number of slots. This not only necessitates prior knowledge of the dataset but also overlooks the inherent variability in the number of objects present in each instance. To overcome this fundamental limitation, we present a novel complexity-aware object auto-encoder framework. Within this framework, we introduce an adaptive slot attention (AdaSlot) mecha-nism that dynamically determines the optimal number of slots based on the content of the data. This is achieved by proposing a discrete slot sampling module that is responsible for selecting an appropriate number of slots from a candidate list. Furthermore, we introduce a masked slot decoder that suppresses unselected slots during the decoding process. Our framework, tested extensively on object discovery tasks with various datasets, shows performance matching or exceeding top fixed-slot models. Moreover, our analysis substantiates that our method exhibits the capability to dynamically adapt the slot number according to each instance's complexity, offering the potential for further exploration in slot attention research. Project will be available at https://kfan21.github.io/AdaSlot/}, author = {Fan, Ke and Bai, Zechen and Xiao, Tianjun and He, Tong and Horn, Max and Fu, Yanwei and Locatello, Francesco and Zhang, Zheng}, booktitle = {2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition}, location = {Seattle, WA, United States}, publisher = {IEEE}, title = {{Adaptive slot attention: Object discovery with dynamic slot number}}, doi = {10.1109/cvpr52733.2024.02176}, year = {2024}, } @inproceedings{14946, abstract = {We present a unified framework for studying the identifiability of representations learned from simultaneously observed views, such as different data modalities. We allow a partially observed setting in which each view constitutes a nonlinear mixture of a subset of underlying latent variables, which can be causally related. We prove that the information shared across all subsets of any number of views can be learned up to a smooth bijection using contrastive learning and a single encoder per view. We also provide graphical criteria indicating which latent variables can be identified through a simple set of rules, which we refer to as identifiability algebra. Our general framework and theoretical results unify and extend several previous work on multi-view nonlinear ICA, disentanglement, and causal representation learning. We experimentally validate our claims on numerical, image, and multi-modal data sets. Further, we demonstrate that the performance of prior methods is recovered in different special cases of our setup. Overall, we find that access to multiple partial views offers unique opportunities for identifiable representation learning, enabling the discovery of latent structures from purely observational data.}, author = {Yao, Dingling and Xu, Danru and Lachapelle, Sébastien and Magliacane, Sara and Taslakian, Perouz and Martius, Georg and Kügelgen, Julius von and Locatello, Francesco}, booktitle = {12th International Conference on Learning Representations}, location = {Vienna, Austria}, publisher = {Curran Associates}, title = {{Multi-view causal representation learning with partial observability}}, year = {2024}, } @inproceedings{18971, abstract = {Models prone to spurious correlations in training data often produce brittle predictions and introduce unintended biases. Addressing this challenge typically involves methods relying on prior knowledge and group annotation to remove spurious correlations, which may not be readily available in many applications. In this paper, we establish a novel connection between unsupervised object-centric learning and mitigation of spurious correlations. Instead of directly inferring subgroups with varying correlations with labels, our approach focuses on discovering concepts: discrete ideas that are shared across input samples. Leveraging existing object-centric representation learning, we introduce CoBalT: a concept balancing technique that effectively mitigates spurious correlations without requiring human labeling of subgroups. Evaluation across the benchmark datasets for sub-population shifts demonstrate superior or competitive performance compared state-of-the-art baselines, without the need for group annotation. Code is available at https://github.com/rarefin/CoBalT}, author = {Arefin, Rifat and Zhang, Yan and Baratin, Aristide and Locatello, Francesco and Rish, Irina and Liu, Dianbo and Kawaguchi, Kenji}, booktitle = {Proceedings of the 41st International Conference on Machine Learning}, issn = {2640-3498}, location = {Vienna, Austria}, pages = {1672--1688}, publisher = {ML Research Press}, title = {{Unsupervised concept discovery mitigates spurious correlations}}, volume = {235}, year = {2024}, }