{"total":461,"rows":[{"id":"2311.03658","title":"The Linear Representation Hypothesis and the Geometry of Large Language Models","created":"2023-11-07","cited_by_count":716,"area":null},{"id":"2106.04619","title":"Self-Supervised Learning with Data Augmentations Provably Isolates Content from Style","created":"2021-06-08","cited_by_count":312,"area":null},{"id":"2304.12210","title":"A Cookbook of Self-Supervised Learning","created":"2023-04-24","cited_by_count":250,"area":null},{"id":"2102.08850","title":"Contrastive Learning Inverts the Data Generating Process","created":"2021-02-17","cited_by_count":221,"area":null},{"id":"2106.06607","title":"Invariance Principle Meets Information Bottleneck for Out-of-Distribution Generalization","created":"2021-06-11","cited_by_count":220,"area":null},{"id":"2107.10098","title":"Disentanglement via Mechanism Sparsity Regularization: A New Principle for Nonlinear ICA","created":"2021-07-21","cited_by_count":160,"area":null},{"id":"2209.11924","title":"Interventional Causal Representation Learning","created":"2022-09-24","cited_by_count":142,"area":null},{"id":"2203.16437","title":"Weakly supervised causal representation learning","created":"2022-03-30","cited_by_count":140,"area":null},{"id":"2202.03169","title":"CITRIS: Causal Identifiability from Temporal Intervened Sequences","created":"2022-02-07","cited_by_count":133,"area":null},{"id":"2211.11695","title":"Disentangled Representation Learning","created":"2022-11-21","cited_by_count":112,"area":null},{"id":"2306.00542","title":"Nonparametric Identifiability of Causal Representations from Unknown Interventions","created":"2023-06-01","cited_by_count":106,"area":null},{"id":"2110.05428","title":"Learning Temporally Causal Latent Processes from General Temporal Data","created":"2021-10-11","cited_by_count":98,"area":null},{"id":"2106.05200","title":"Independent mechanism analysis, a new concept?","created":"2021-06-09","cited_by_count":97,"area":null},{"id":"2306.02235","title":"Learning Linear Causal Representations from Interventions under General Nonlinear Mixing","created":"2023-06-04","cited_by_count":92,"area":null},{"id":"2206.12126","title":"Temporal Attention Unit: Towards Efficient Spatiotemporal Predictive Learning","created":"2022-06-24","cited_by_count":86,"area":null},{"id":"2206.07751","title":"On the Identifiability of Nonlinear ICA: Sparsity and Beyond","created":"2022-06-15","cited_by_count":85,"area":null},{"id":"2211.16467","title":"Linear Causal Disentanglement via Interventions","created":"2022-11-29","cited_by_count":84,"area":null},{"id":"2307.06250","title":"Identifiability Guarantees for Causal Disentanglement from Soft Interventions","created":"2023-07-12","cited_by_count":84,"area":null},{"id":"2403.03867","title":"On the Origins of Linear Representations in Large Language Models","created":"2024-03-06","cited_by_count":79,"area":null},{"id":"2109.03795","title":"Desiderata for Representation Learning: A Causal Perspective","created":"2021-09-08","cited_by_count":76,"area":null},{"id":"2205.15612","title":"GlanceNets: Interpretabile, Leak-proof Concept-based Models","created":"2022-05-31","cited_by_count":73,"area":null},{"id":"2111.03042","title":"Unsupervised Learning of Compositional Energy Concepts","created":"2021-11-04","cited_by_count":71,"area":null},{"id":"2206.10044","title":"Identifiability of deep generative models without auxiliary information","created":"2022-06-20","cited_by_count":70,"area":null},{"id":"2402.05052","title":"Causal Representation Learning from Multiple Distributions: A General Setting","created":"2024-02-07","cited_by_count":68,"area":null},{"id":"2206.01101","title":"Weakly Supervised Representation Learning with Sparse Perturbations","created":"2022-06-02","cited_by_count":65,"area":null},{"id":"2106.09620","title":"Disentangling Identifiable Features from Noisy Data with Structured Nonlinear ICA","created":"2021-06-17","cited_by_count":64,"area":null},{"id":"2305.19951","title":"Not All Neuro-Symbolic Concepts Are Created Equal: Analysis and Mitigation of Reasoning Shortcuts","created":"2023-05-31","cited_by_count":63,"area":null},{"id":"2107.08221","title":"Visual Representation Learning Does Not Generalize Strongly Within the Same Domain","created":"2021-07-17","cited_by_count":61,"area":null},{"id":"2210.13647","title":"Temporally Disentangled Representation Learning","created":"2022-10-24","cited_by_count":61,"area":null},{"id":"2303.09166","title":"Identifiability Results for Multimodal Contrastive Learning","created":"2023-03-16","cited_by_count":61,"area":null},{"id":"2203.05818","title":"ZIN: When and How to Learn Invariance Without Environment Partition?","created":"2022-03-11","cited_by_count":58,"area":null},{"id":"2302.02672","title":"Identifiability of latent-variable and structural-equation models: from linear to nonlinear","created":"2023-02-06","cited_by_count":58,"area":null},{"id":"2307.02598","title":"Additive Decoders for Latent Variables Identification and Cartesian-Product Extrapolation","created":"2023-07-05","cited_by_count":58,"area":null},{"id":"2301.00537","title":"Posterior Collapse and Latent Variable Non-identifiability","created":"2023-01-02","cited_by_count":57,"area":null},{"id":"2103.10226","title":"Beyond Trivial Counterfactual Explanations with Diverse Valuable Explanations","created":"2021-03-18","cited_by_count":56,"area":null},{"id":"2305.17225","title":"Causal Component Analysis","created":"2023-05-26","cited_by_count":56,"area":null},{"id":"2106.15563","title":"Learning latent causal graphs via mixture oracles","created":"2021-06-29","cited_by_count":55,"area":null},{"id":"2110.10804","title":"Identifiable Deep Generative Models via Sparse Decoding","created":"2021-10-20","cited_by_count":54,"area":null},{"id":"2208.06406","title":"Function Classes for Identifiable Nonlinear Independent Component Analysis","created":"2022-08-12","cited_by_count":53,"area":null},{"id":"2306.15764","title":"High Fidelity Image Counterfactuals with Probabilistic Causal Models","created":"2023-06-27","cited_by_count":53,"area":null},{"id":"2206.06169","title":"Causal Representation Learning for Instantaneous and Temporal Effects in Interactive Systems","created":"2022-06-13","cited_by_count":51,"area":null},{"id":"2211.08422","title":"Mechanistic Mode Connectivity","created":"2022-11-15","cited_by_count":50,"area":null},{"id":"2205.00403","title":"A Simple Approach to Improve Single-Model Deep Uncertainty via Distance-Awareness","created":"2022-05-01","cited_by_count":49,"area":null},{"id":"2301.08230","title":"Score-based Causal Representation Learning with Interventions","created":"2023-01-19","cited_by_count":49,"area":null},{"id":"2210.08443","title":"CLEAR: Generative Counterfactual Explanations on Graphs","created":"2022-10-16","cited_by_count":48,"area":null},{"id":"2303.16535","title":"Nonlinear Independent Component Analysis for Principled Disentanglement in Unsupervised Deep Learning","created":"2023-03-29","cited_by_count":48,"area":null},{"id":"2108.12043","title":"Learning Disentangled Representations in the Imaging Domain","created":"2021-08-26","cited_by_count":46,"area":null},{"id":"2305.14229","title":"Provably Learning Object-Centric Representations","created":"2023-05-23","cited_by_count":45,"area":null},{"id":"2211.14666","title":"Synergies between Disentanglement and Sparsity: Generalization and Identifiability in Multi-Task Learning","created":"2022-11-26","cited_by_count":44,"area":null},{"id":"2302.02228","title":"Counterfactual Identifiability of Bijective Causal Models","created":"2023-02-04","cited_by_count":43,"area":null},{"id":"2303.10112","title":"Causal Discovery from Temporal Data: An Overview and New Perspectives","created":"2023-03-17","cited_by_count":43,"area":null},{"id":"2102.12353","title":"Nonlinear Invariant Risk Minimization: A Causal Approach","created":"2021-02-24","cited_by_count":42,"area":null},{"id":"2310.05327","title":"Provable Compositional Generalization for Object-Centric Learning","created":"2023-10-09","cited_by_count":41,"area":null},{"id":"2305.18378","title":"Disentanglement via Latent Quantization","created":"2023-05-28","cited_by_count":40,"area":null},{"id":"2306.04898","title":"Understanding Masked Autoencoders via Hierarchical Latent Variable Models","created":"2023-06-08","cited_by_count":39,"area":null},{"id":"2306.09643","title":"BISCUIT: Causal Representation Learning from Binary Interactions","created":"2023-06-16","cited_by_count":38,"area":null},{"id":"2106.07115","title":"Understanding Latent Correlation-Based Multiview Learning and Self-Supervision: An Identifiability Perspective","created":"2021-06-14","cited_by_count":37,"area":null},{"id":"2210.07347","title":"Disentanglement of Correlated Factors via Hausdorff Factorized Support","created":"2022-10-13","cited_by_count":37,"area":null},{"id":"2306.11078","title":"Beyond Normal: On the Evaluation of Mutual Information Estimators","created":"2023-06-19","cited_by_count":37,"area":null},{"id":"2310.04723","title":"Subspace Identification for Multi-Source Domain Adaptation","created":"2023-10-07","cited_by_count":37,"area":null},{"id":"2402.12240","title":"BEARS Make Neuro-Symbolic Models Aware of their Reasoning Shortcuts","created":"2024-02-19","cited_by_count":37,"area":null},{"id":"2104.14235","title":"Inspect, Understand, Overcome: A Survey of Practical Methods for AI Safety","created":"2021-04-29","cited_by_count":35,"area":null},{"id":"2403.08335","title":"A Sparsity Principle for Partially Observable Causal Representation Learning","created":"2024-03-13","cited_by_count":35,"area":null},{"id":"2110.12091","title":"Contrastively Disentangled Sequential Variational Autoencoder","created":"2021-10-22","cited_by_count":34,"area":null},{"id":"2208.14153","title":"Identifying Weight-Variant Latent Causal Models","created":"2022-08-30","cited_by_count":34,"area":null},{"id":"2211.03553","title":"Learning Causal Representations of Single Cells via Sparse Mechanism Shift Modeling","created":"2022-11-07","cited_by_count":34,"area":null},{"id":"2303.01274","title":"Measuring axiomatic soundness of counterfactual image models","created":"2023-03-02","cited_by_count":34,"area":null},{"id":"2306.02899","title":"Learning nonparametric latent causal graphs with unknown interventions","created":"2023-06-05","cited_by_count":34,"area":null},{"id":"2310.15450","title":"General Identifiability and Achievability for Causal Representation Learning","created":"2023-10-24","cited_by_count":34,"area":null},{"id":"2311.00866","title":"Generalizing Nonlinear ICA Beyond Structural Sparsity","created":"2023-11-01","cited_by_count":34,"area":null},{"id":"2206.00801","title":"Indeterminacy in Generative Models: Characterization and Strong Identifiability","created":"2022-06-02","cited_by_count":33,"area":null},{"id":"2302.00993","title":"Unpaired Multi-Domain Causal Representation Learning","created":"2023-02-02","cited_by_count":32,"area":null},{"id":"2304.07939","title":"Leveraging sparse and shared feature activations for disentangled representation learning","created":"2023-04-17","cited_by_count":32,"area":null},{"id":"2310.15709","title":"Causal Representation Learning Made Identifiable by Grouping of Observational Variables","created":"2023-10-24","cited_by_count":32,"area":null},{"id":"2410.21869","title":"Cross-Entropy Is All You Need To Invert the Data Generating Process","created":"2024-10-29","cited_by_count":31,"area":null},{"id":"2106.09913","title":"Iterative Feature Matching: Toward Provable Domain Generalization with Logarithmic Environments","created":"2021-06-18","cited_by_count":30,"area":null},{"id":"2110.14708","title":"Identifiable Generative Models for Missing Not at Random Data Imputation","created":"2021-10-27","cited_by_count":30,"area":null},{"id":"2110.15796","title":"Properties from Mechanisms: An Equivariance Perspective on Identifiable Representation Learning","created":"2021-10-29","cited_by_count":30,"area":null},{"id":"2210.01768","title":"Disentanglement with Biological Constraints: A Theory of Functional Cell Types","created":"2022-09-30","cited_by_count":30,"area":null},{"id":"2207.07732","title":"Partial Disentanglement via Mechanism Sparsity","created":"2022-07-15","cited_by_count":29,"area":null},{"id":"2402.09236","title":"Learning Interpretable Concepts: Unifying Causal Representation Learning and Foundation Models","created":"2024-02-14","cited_by_count":29,"area":null},{"id":"2206.02416","title":"Embrace the Gap: VAEs Perform Independent Mechanism Analysis","created":"2022-06-06","cited_by_count":28,"area":null},{"id":"2306.07916","title":"Identification of Nonlinear Latent Hierarchical Models","created":"2023-06-13","cited_by_count":27,"area":null},{"id":"2207.02598","title":"Predicting is not Understanding: Recognizing and Addressing Underspecification in Machine Learning","created":"2022-07-06","cited_by_count":26,"area":null},{"id":"2210.00364","title":"DCI-ES: An Extended Disentanglement Framework with Connections to Identifiability","created":"2022-10-01","cited_by_count":23,"area":null},{"id":"2310.02854","title":"Multi-Domain Causal Representation Learning via Weak Distributional Invariances","created":"2023-10-04","cited_by_count":23,"area":null},{"id":"2310.15580","title":"Identifiable Latent Polynomial Causal Models Through the Lens of Change","created":"2023-10-24","cited_by_count":23,"area":null},{"id":"2401.14535","title":"CaRiNG: Learning Temporal Causal Representation under Non-Invertible Generation Process","created":"2024-01-25","cited_by_count":23,"area":null},{"id":"2102.10303","title":"Towards Building A Group-based Unsupervised Representation Disentanglement Framework","created":"2021-02-20","cited_by_count":22,"area":null},{"id":"2309.07742","title":"Interpretability is in the Mind of the Beholder: A Causal Framework for Human-interpretable Representation Learning","created":"2023-09-14","cited_by_count":22,"area":null},{"id":"2310.11011","title":"From Identifiable Causal Representations to Controllable Counterfactual Generation: A Survey on Causal Generative Modeling","created":"2023-10-17","cited_by_count":22,"area":null},{"id":"2310.18615","title":"Temporally Disentangled Representation Learning under Unknown Nonstationarity","created":"2023-10-28","cited_by_count":22,"area":null},{"id":"2402.04875","title":"On Provable Length and Compositional Generalization","created":"2024-02-07","cited_by_count":22,"area":null},{"id":"2106.05238","title":"I Don't Need u: Identifiable Non-Linear ICA Without Side Information","created":"2021-06-09","cited_by_count":21,"area":null},{"id":"2310.04295","title":"Identifying Representations for Intervention Extrapolation","created":"2023-10-06","cited_by_count":21,"area":null},{"id":"2310.19054","title":"Object-centric architectures enable efficient causal representation learning","created":"2023-10-29","cited_by_count":21,"area":null},{"id":"2206.05263","title":"Causal Balancing for Domain Generalization","created":"2022-06-10","cited_by_count":20,"area":null},{"id":"2303.03052","title":"Masked Images Are Counterfactual Samples for Robust Fine-tuning","created":"2023-03-06","cited_by_count":20,"area":null},{"id":"2403.09683","title":"Counterfactual Image Editing","created":"2024-02-07","cited_by_count":20,"area":null},{"id":"2411.07784","title":"Interaction Asymmetry: A General Principle for Learning Composable Abstractions","created":"2024-11-12","cited_by_count":20,"area":null}],"next_cursor":"100"}