Dhanya Sridhar
Assistant Professor
Contact
Dhanya Sridhar
Assistant Professor
I'm an assistant professor in the department of Informatics and Operations Research (DIRO) at Université de Montréal and a core academic member of Mila - Quebec Artificial Intelligence Institute. I'm fortunate to receive support as a Canada CIFAR AI Chair. I also co-lead the IVADO R3AI working group on safe and aligned AI.
In brief, my research focuses on combining causality and machine learning in service of AI systems that are robust to distribution shifts, adapt to new tasks efficiently, and discover new knowledge alongside us. This large vision includes learning causal representations to interpret complex and unstructured data with limited human supervision, predictors that learn causal mechanisms to remain robust, interpreting large AI systems with causal abstraction, understanding and improving new learning paradigms like in-context learning, and aspects of responsible AI.
In brief, my research focuses on combining causality and machine learning in service of AI systems that are robust to distribution shifts, adapt to new tasks efficiently, and discover new knowledge alongside us. This large vision includes learning causal representations to interpret complex and unstructured data with limited human supervision, predictors that learn causal mechanisms to remain robust, interpreting large AI systems with causal abstraction, understanding and improving new learning paradigms like in-context learning, and aspects of responsible AI.
Research group:
PhD Students
Philippe Brouillard, co-supervised with Alexandre Drouin
Shruti Joshi
Mizu Nishikawa-Toomey, co-supervised with Laurent Charlin
Tom Marty
Cristian Manta, co-supervised with Yoshua Bengio
Sophia Gunluk
Philippe Brouillard, co-supervised with Alexandre Drouin
Shruti Joshi
Mizu Nishikawa-Toomey, co-supervised with Laurent Charlin
Tom Marty
Cristian Manta, co-supervised with Yoshua Bengio
Sophia Gunluk
Former interns
Navita Goyal, PhD student, University of Maryland, College Park
Maitreyi Swaroop, MSc student, Indian Institute of Technology, Kharagpur
Navita Goyal, PhD student, University of Maryland, College Park
Maitreyi Swaroop, MSc student, Indian Institute of Technology, Kharagpur
Selected Publications
Sparsity regularization via tree-structured environments for disentangled representations
Elliot Layne, Dhanya Sridhar, Jason S. Hartford, M. Blanchette
2024
Demystifying amortized causal discovery with transformers
Francesco Montagna, Max Cairney-Leeming, Dhanya Sridhar, Francesco Locatello
arXiv.org, 2024
Causal Representation Learning in Temporal Data via Single-Parent Decoding
Philippe Brouillard, Sébastien Lachapelle, Julia Kaltenborn, Yaniv Gurwicz, Dhanya Sridhar, Alexandre Drouin, Peer Nowack, Jakob Runge, David Rolnick
2024
Does learning the right latent variables necessarily improve in-context learning?
Sarthak Mittal, Eric Elmoznino, L'eo Gagnon, Sangnie Bhardwaj, Dhanya Sridhar, Guillaume Lajoie
arXiv.org, 2024
In-Context Learning Can Re-learn Forbidden Tasks
Sophie Xhonneux, David Dobre, Jian Tang, Gauthier Gidel, Dhanya Sridhar
arXiv.org, 2024
View all
Courses
Causal Inference and ML
This course combines lectures and seminar-style discussions to cover the foundations of causality and topics like causal representation learning, causal structure discovery, causal abstraction (and its use in understanding large models).
Causal Inference and ML
This course combines lectures and seminar-style discussions to cover the foundations of causality and topics like causal representation learning, causal structure discovery, causal abstraction (and its use in understanding large models).
Fundamentals of Machine Learning