paper-with-me

Papers

Learning Linear Causal Representations from Interventions under General Nonlinear Mixing

2023-06-04 · NeurIPS 2023 11 · Simon Buchholz, Goutham Rajendran, Elan Rosenfeld, Bryon Aragam, Bernhard Schölkopf, Pradeep Ravikumar

We study the problem of learning causal representations from unknown, latent interventions in a general setting, where the latent distribution is Gaussian but the mixing function is completely general. We prove strong identifiability results given unknown single-node interventions, i.e., without having access to the intervention targets. This generalizes prior works which have focused on weaker classes, such as linear maps or paired counterfactual data. This is also the first instance of causal identifiability from non-paired interventions for deep neural network embeddings. Our proof relies on carefully uncovering the high-dimensional geometric structure present in the data distribution after a non-linear density transformation, which we capture by analyzing quadratic forms of precision matrices of the latent distributions. Finally, we propose a contrastive algorithm to identify the latent variables in practice and evaluate its performance on various tasks.

📄 PDF Abstract BibTeX arXiv:2306.02235

Code (1)

simonbing/crlsanitycheck pytorch

Tasks

counterfactual

Similar Papers 제목 키워드 기반

Learning nonparametric latent causal graphs with unknown interventions

2023-06-05 · NeurIPS 2023 11

We establish conditions under which latent causal graphs are nonparametrically identifiable and can be reconstructed from unknown interventions in the latent space. Our primary focus is the identification of the latent s…

Identifying Linearly-Mixed Causal Representations from Multi-Node Interventions

2023-11-05 · Simon Bing, Urmi Ninad, jonas Wahl, Jakob Runge

The task of inferring high-level causal variables from low-level observations, commonly referred to as causal representation learning, is fundamentally underconstrained. As such, recent works to address this problem focu…

Representation Learning

Beyond identifiability: Learning causal representations with few environments and finite samples

2026-03-26 · Inbeom Lee, Tongtong Jin, Bryon Aragam arxiv

We provide explicit, finite-sample guarantees for learning causal representations from data with a sublinear number of environments. Causal representation learning seeks to provide a rigourous foundation for the general …

Representation Learning

Learning Causal Representations from General Environments: Identifiability and Intrinsic Ambiguity

2023-11-21 · Jikai Jin, Vasilis Syrgkanis

We study causal representation learning, the task of recovering high-level latent variables and their causal relationships in the form of a causal graph from low-level observed data (such as text and images), assuming ac…

Representation Learning

Nonparametric Identifiability of Causal Representations from Unknown Interventions

2023-06-01 · NeurIPS 2023 11 · Julius von Kügelgen, Michel Besserve, Liang Wendong, Luigi Gresele 외

We study causal representation learning, the task of inferring latent causal variables and their causal relations from high-dimensional mixtures of the variables. Prior work relies on weak supervision, in the form of cou…

counterfactualRepresentation Learning