Conceptual Challenges In Connecting Interpretability And Causality

Channel:
United States Simons Institute for the Theory of Computing
Subscribers:
68,700
Published on ● Video Link: https://www.youtube.com/watch?v=9UvgiQAoyb0


Duration: 38:40
813 views
17

Alex D'Amour (Google Brain)
https://simons.berkeley.edu/talks/conceptual-challenges-connecting-interpretability-and-causality
Interpretable Machine Learning in Natural and Social Sciences

There has been a strong intuition in the Machine Learning community that interpretability and causality ought to have a strong connection. However, the community has not arrived at consensus about how to formalize this connection. In this talk, I will raise questions about conceptual and technical ambiguities that I think make this connection hard to specify. The goal of the talk is to raise points for discussion, expressed in causal formalism, rather than to provide answers.



Other Videos By Simons Institute for the Theory of Computing


2022-07-01Legal Barriers To Interpretable Machine Learning
2022-06-30Interpretability and Algorithmic Fairness
2022-06-30Panel on Interpretability in the Biological Sciences
2022-06-30Machine Learning, Deep Networks and Interpretability in Systems, Cognitive and...
2022-06-30Interpreting Deep Learning Models Of Functional Genomics Data To Decode Regulatory Sequence...
2022-06-30Panel on Interpretability in the Physical Sciences
2022-06-30Interpreting Machine Learning From the Perspective of Nonequilibrium Systems
2022-06-29Interpretability In Atomic-Scale Machine Learning
2022-06-29Panel on Causality
2022-06-29Explanation: A(N Abridged) Survey
2022-06-29Conceptual Challenges In Connecting Interpretability And Causality
2022-06-29Panel on Societal Dimensions of Explanation
2022-06-29Accuracy and Interpretability Through the Lens of Human-AI Teaming
2022-06-28Goals And Interpretable Variables In Neuroscience
2022-06-18Continuous LWE is as Hard as LWE
2022-06-18Batch Arguments for NP and SNARGs for P from LWE
2022-06-18Lattice-based Zero-knowledge Proofs
2022-06-18Candidate Witness Encryption from Lattice Techniques
2022-06-17Optimal Broadcast Encryption and More from Evasive LWE
2022-06-17Single-Server Private Information Retrieval with Sublinear Amortized Time
2022-06-17Doubly Efficient Private Information Retrieval


Tags:
Simons Institute
theoretical computer science
UC Berkeley
Computer Science
Theory of Computation
Theory of Computing
Interpretable Machine Learning in Natural and Social Sciences
Alex D'Amour