paper-with-me

홈 › Papers

A Meta-Transfer Objective for Learning to Disentangle Causal Mechanisms

2019-01-30 · ICLR 2020 1 · Yoshua Bengio, Tristan Deleu, Nasim Rahaman, Rosemary Ke, Sébastien Lachapelle, Olexa Bilaniuk, Anirudh Goyal, Christopher Pal

We propose to meta-learn causal structures based on how fast a learner adapts to new distributions arising from sparse distributional changes, e.g. due to interventions, actions of agents and other sources of non-stationarities. We show that under this assumption, the correct causal structural choices lead to faster adaptation to modified distributions because the changes are concentrated in one or just a few mechanisms when the learned knowledge is modularized appropriately. This leads to sparse expected gradients and a lower effective number of degrees of freedom needing to be relearned while adapting to the change. It motivates using the speed of adaptation to a modified distribution as a meta-learning objective. We demonstrate how this can be used to determine the cause-effect relationship between two observed variables. The distributional changes do not need to correspond to standard interventions (clamping a variable), and the learner has no direct knowledge of these interventions. We show that causal structures can be parameterized via continuous variables and learned end-to-end. We then explore how these ideas could be used to also learn an encoder that would map low-level observed variables to unobserved causal variables leading to faster adaptation out-of-distribution, learning a representation space where one can satisfy the assumptions of independent mechanisms and of small and sparse changes in these mechanisms due to actions and non-stationarities.

📄 PDF Abstract BibTeX arXiv:1901.10912

Code (2)

authors-1901-10912/A-Meta-Transfer-Objective-For-Learning-To-Disentangle-Causal-Mechanisms 공식 구현 pytorch
ec6dde01667145e58de60f864e05a4/CausalOptimizationAnon 공식 구현 pytorch

Tasks

Meta-Learning

Methods 이 논문이 사용한 방법론

SPEED The monocular depth estimation (MDE) is the task of estimating depth from a single frame. This information is an essential knowledge in many computer vision tasks such as scene…

Similar Papers 제목 키워드 기반

Connecting Data to Mechanisms with Meta Structual Causal Model

2021-09-29 · Gong Heyang

Recent years have seen impressive progress in theoretical and algorithmic developments of causal inference across various disciplines in science and engineering. However, there is still some unresolved theoretical proble…

Causal InferenceRepresentation Learning

Learning Causally Disentangled Representations via the Principle of Independent Causal Mechanisms

2023-06-02 · Aneesh Komanduri, Yongkai Wu, Feng Chen, Xintao Wu

Learning disentangled causal representations is a challenging problem that has gained significant attention recently due to its implications for extracting meaningful information for downstream tasks. In this work, we de…

counterfactualDisentanglement

Bayesian Meta-Learning with Expert Feedback for Task-Shift Adaptation through Causal Embeddings

2026-02-23 · Lotta Mäkinen, Jorge Loría, Samuel Kaski arxiv

Meta-learning methods perform well on new within-distribution tasks but often fail when adapting to out-of-distribution target tasks, where transfer from source tasks can induce negative transfer. We propose a causally-a…

Concept-free Causal Disentanglement with Variational Graph Auto-Encoder

2023-11-17 · Jingyun Feng, Lin Zhang, Lili Yang

In disentangled representation learning, the goal is to achieve a compact representation that consists of all interpretable generative factors in the observational data. Learning disentangled representations for graphs b…

DisentanglementMeta-LearningRepresentation Learning

Causal-Symbolic Meta-Learning (CSML): Inducing Causal World Models for Few-Shot Generalization

2025-09-15 · Mohamed Zayaan S arxiv

Modern deep learning models excel at pattern recognition but remain fundamentally limited by their reliance on spurious correlations, leading to poor generalization and a demand for massive datasets. We argue that a key …

Causal Inference