Toward Causal Representation Learning

Schölkopf, Bernhard; Locatello, Francesco; Bauer, Stefan; Ke, Nan Rosemary; Kalchbrenner, Nal; Goyal, Anirudh; Bengio, Yoshua

doi:10.1109/jproc.2021.3058954

Cited by 614 publications

(430 citation statements)

References 103 publications

(136 reference statements)

Supporting

Mentioning

426

Contrasting

Unclassified

Order By: Relevance

“…[69][70][71][72][73] However, understanding and disentangling their latent spaces remains challenging. 74,75 viAE wins Among the tested DNNs and across the multiple tests, the viAE provided the best face-shape representations to predict human behavior. With the notable exception of the generalization testing, the simple nonlinear pixelPCA model came close to this performance.…”

Section: Hypothesis-driven Research Using Generative Modelsmentioning

confidence: 96%

Grounding deep neural network predictions of human categorization behavior in understandable functional features: The case of face identity

Daube

Zhan

et al. 2021

Patterns

View full text Add to dashboard Cite

show abstract

Section: Hypothesis-driven Research Using Generative Modelsmentioning

confidence: 96%

Grounding deep neural network predictions of human categorization behavior in understandable functional features: The case of face identity

Daube

Zhan

et al. 2021

Patterns

View full text Add to dashboard Cite

show abstract

“…However, this teaching has been criticised as being detrimental to the potential understanding, which can be gained from techniques such as counterfactual explanations, a specific class of explanation that provides a link between what could have happened had input to a model been changed in a particular way [ 129 ]. Causal representation learning is a by-product of this research activity, and its applications have reached explainable CD [ 130 , 131 ].…”

Section: Challenges Comparisons and Future Directions For Change Representation Techniquesmentioning

confidence: 99%

Representation Learning for Fine-Grained Change Detection

Mahony

Campbell

Krpálková

et al. 2021

Sensors

View full text Add to dashboard Cite

Fine-grained change detection in sensor data is very challenging for artificial intelligence though it is critically important in practice. It is the process of identifying differences in the state of an object or phenomenon where the differences are class-specific and are difficult to generalise. As a result, many recent technologies that leverage big data and deep learning struggle with this task. This review focuses on the state-of-the-art methods, applications, and challenges of representation learning for fine-grained change detection. Our research focuses on methods of harnessing the latent metric space of representation learning techniques as an interim output for hybrid human-machine intelligence. We review methods for transforming and projecting embedding space such that significant changes can be communicated more effectively and a more comprehensive interpretation of underlying relationships in sensor data is facilitated. We conduct this research in our work towards developing a method for aligning the axes of latent embedding space with meaningful real-world metrics so that the reasoning behind the detection of change in relation to past observations may be revealed and adjusted. This is an important topic in many fields concerned with producing more meaningful and explainable outputs from deep learning and also for providing means for knowledge injection and model calibration in order to maintain user confidence.

show abstract

“…This is the essential problem for artificial systems in emulating cognition in animals. However, there is recent work that employs artificial models of transfer learning [21,22].…”

Section: Abstract Encoding Of Sensory Inputmentioning

confidence: 99%

Cognition as a Mechanical Process

Friedman

2021

NeuroSci

View full text Add to dashboard Cite

Cognition is often defined as a dual process of physical and non-physical mechanisms. This duality originated from past theory on the constituent parts of the natural world. Even though material causation is not an explanation for all natural processes, phenomena at the cellular level of life are modeled by physical causes. These phenomena include explanations for the function of organ systems, including the nervous system and information processing in the cerebrum. This review restricts the definition of cognition to a mechanistic process and enlists studies that support an abstract set of proximate mechanisms. Specifically, this process is approached from a large-scale perspective, the flow of information in a neural system. Study at this scale further constrains the possible explanations for cognition since the information flow is amenable to theory, unlike a lower-level approach where the problem becomes intractable. These possible hypotheses include stochastic processes for explaining the processes of cognition along with principles that support an abstract format for the encoded information.

show abstract

Toward Causal Representation Learning

Cited by 614 publications

References 103 publications

Grounding deep neural network predictions of human categorization behavior in understandable functional features: The case of face identity

Grounding deep neural network predictions of human categorization behavior in understandable functional features: The case of face identity

Representation Learning for Fine-Grained Change Detection

Cognition as a Mechanical Process

Contact Info

Product

Resources

About