Workshop
Workshop on neuro Causal and Symbolic AI (nCSI)
Matej Zečević · Devendra Dhami · Christina Winkler · Thomas Kipf · Robert Peharz · Petar Veličković
Virtual
Fri 9 Dec, 4 a.m. PST
Understanding causal interactions is central to human cognition and thereby a central quest in science, engineering, business, and law. Developmental psychology has shown that children explore the world in a similar way to how scientists do, asking questions such as “What if?” and “Why?” AI research aims to replicate these capabilities in machines. Deep learning in particular has brought about powerful tools for function approximation by means of end-to-end traininable deep neural networks. This capability has been corroborated by tremendous success in countless applications. However, their lack of interpretability and reasoning capabilities prove to be a hindrance towards building systems of human-like ability. Therefore, enabling causal reasoning capabilities in deep learning is of critical importance for research on the path towards human-level intelligence. First steps towards neural-causal models exist and promise a vision of AI systems that perform causal inferences as efficiently as modern-day neural models. Similarly, classical symbolic methods are being revisited and reintegrated into current systems to allow for reasoning capabilities beyond pure pattern recognition. The Pearlian formalization to causality has revealed a theoretically sound and practically strict hierarchy of reasoning that serves as a helpful benchmark for evaluating the reasoning capabilities of neuro-symbolic systems.
Our aim is to bring together researchers interested in the integration of research areas in artificial intelligence (general machine and deep learning, symbolic and object-centric methods, and logic) with rigorous formalizations of causality with the goal of developing next-generation AI systems.
Schedule
Fri 4:00 a.m. - 4:10 a.m.
|
Welcome & Opening Remarks
(
Opening
)
>
SlidesLive Video |
Matej Zečević 🔗 |
Fri 4:10 a.m. - 4:40 a.m.
|
Opening Keynote for nCSI
(
Invited Speaker
)
>
SlidesLive Video |
Judea Pearl 🔗 |
Fri 4:40 a.m. - 4:50 a.m.
|
GlanceNets: Interpretable, Leak-proof Concept-based Models
(
Oral
)
>
link
SlidesLive Video |
Emanuele Marconato · Andrea Passerini · Stefano Teso 🔗 |
Fri 4:50 a.m. - 5:00 a.m.
|
Meaning without reference in large language models
(
Oral
)
>
link
SlidesLive Video |
Steven Piantadosi · Felix Hill 🔗 |
Fri 5:00 a.m. - 6:30 a.m.
|
Virtual Poster Session 1 ( Poster Session ) > link | 🔗 |
Fri 6:30 a.m. - 7:00 a.m.
|
Neural Models with Symbolic Representations for Perceptuo-Reasoning Tasks
(
Invited Speaker
)
>
SlidesLive Video |
- Mausam 🔗 |
Fri 7:00 a.m. - 7:30 a.m.
|
Causal Inference from Text
(
Invited Speaker
)
>
SlidesLive Video |
Dhanya Sridhar 🔗 |
Fri 7:30 a.m. - 8:30 a.m.
|
Break
|
🔗 |
Fri 8:30 a.m. - 8:40 a.m.
|
Unlocking Slot Attention by Changing Optimal Transport Costs
(
Oral
)
>
link
SlidesLive Video |
Yan Zhang · David Zhang · Simon Lacoste-Julien · Gertjan Burghouts · Cees Snoek 🔗 |
Fri 8:40 a.m. - 8:50 a.m.
|
Interventional Causal Representation Learning
(
Oral
)
>
link
SlidesLive Video |
Kartik Ahuja · Yixin Wang · Divyat Mahajan · Yoshua Bengio 🔗 |
Fri 8:50 a.m. - 9:20 a.m.
|
Representation Learning and Causality
(
Invited Speaker
)
>
SlidesLive Video |
Jovana Mitrovic 🔗 |
Fri 9:20 a.m. - 10:30 a.m.
|
Virtual Poster Session 2 ( Poster Session ) > link | 🔗 |
Fri 10:30 a.m. - 11:00 a.m.
|
A Counterfactual Simulation Model of Causal Judgment
(
Invited Speaker
)
>
SlidesLive Video |
Tobias Gerstenberg 🔗 |
Fri 11:00 a.m. - 11:30 a.m.
|
AI can learn from data. But can it learn to reason?
(
Invited Speaker
)
>
SlidesLive Video |
Guy Van den Broeck 🔗 |
Fri 11:30 a.m. - 12:00 p.m.
|
Break
|
🔗 |
Fri 12:00 p.m. - 12:50 p.m.
|
Panel Discussion: "Heading for a Unifying View on nCSI"
(
Panel
)
>
SlidesLive Video |
Tobias Gerstenberg · Sriraam Natarajan · - Mausam · Guy Van den Broeck · Devendra Dhami 🔗 |
Fri 12:50 p.m. - 1:00 p.m.
|
Closing Remarks
(
Closing
)
>
SlidesLive Video |
Matej Zečević 🔗 |
-
|
Synthesized Differentiable Programs ( Poster ) > link | Lucas Saldyt 🔗 |
-
|
Probabilities of Causation: Adequate Size of Experimental and Observational Samples ( Poster ) > link | Ang Li · Ruirui Mao · Judea Pearl 🔗 |
-
|
Discrete Learning Of DAGs Via Backpropagation ( Poster ) > link | Andrew Wren · Pasquale Minervini · Luca Franceschi · Valentina Zantedeschi 🔗 |
-
|
Symbolic Causal Inference via Operations on Probabilistic Circuits ( Poster ) > link | Benjie Wang · Marta Kwiatkowska 🔗 |
-
|
Benchmarking Counterfactual Reasoning Abilities about Implicit Physical Properties ( Poster ) > link | Maitreya Patel · Tejas Gokhale · Chitta Baral · 'YZ' Yezhou Yang 🔗 |
-
|
Hierarchical Abstraction for Combinatorial Generalization in Object Rearrangement ( Poster ) > link | Michael Chang · Alyssa L Dayan · Franziska Meier · Tom Griffiths · Sergey Levine · Amy Zhang 🔗 |
-
|
Enhancing Transfer of Reinforcement Learning Agents with Abstract Contextual Embeddings ( Poster ) > link | Guy Azran · Mohamad Hosein Danesh · Stefano Albrecht · Sarah Keren 🔗 |
-
|
Playgrounds for Abstraction and Reasoning ( Poster ) > link | Subin Kim · Prin Phunyaphibarn · Donghyun Ahn · Sundong Kim 🔗 |
-
|
Causal Discovery for Modular World Models ( Poster ) > link | Anson Lei · Bernhard Schölkopf · Ingmar Posner 🔗 |
-
|
Active Bayesian Causal Inference ( Poster ) > link | Christian Toth · Lars Lorch · Christian Knoll · Andreas Krause · Franz Pernkopf · Robert Peharz · Julius von Kügelgen 🔗 |
-
|
Learning Neuro-symbolic Programs for Language-Guided Robotic Manipulation ( Poster ) > link | Namasivayam Kalithasan · Himanshu Singh · Vishal Bindal · Arnav Tuli · Vishwajeet Agrawal · Rahul Jain · Parag Singla · Rohan Paul 🔗 |
-
|
Image Manipulation via Neuro-Symbolic Networks ( Poster ) > link | Harman Singh · Poorva Garg · Mohit Gupta · Kevin Shah · Arnab Kumar Mondal · Dinesh Khandelwal · Parag Singla · Dinesh Garg 🔗 |
-
|
Counterfactual reasoning: Do Language Models need world knowledge for causal inference? ( Poster ) > link | Jiaxuan Li · Lang Yu · Allyson Ettinger 🔗 |
-
|
Graphs, Constraints, and Search for the Abstraction and Reasoning Corpus ( Poster ) > link | Yudong Xu · Elias Khalil · Scott Sanner 🔗 |
-
|
The Impact of Symbolic Representations on In-context Learning for Few-shot Reasoning ( Poster ) > link | Hanlin Zhang · yifan zhang · Li Erran Li · Eric Xing 🔗 |
-
|
GlanceNets: Interpretable, Leak-proof Concept-based Models ( Poster ) > link | Emanuele Marconato · Andrea Passerini · Stefano Teso 🔗 |
-
|
Interventional Causal Representation Learning ( Poster ) > link | Kartik Ahuja · Yixin Wang · Divyat Mahajan · Yoshua Bengio 🔗 |
-
|
Unlocking Slot Attention by Changing Optimal Transport Costs ( Poster ) > link | Yan Zhang · David Zhang · Simon Lacoste-Julien · Gertjan Burghouts · Cees Snoek 🔗 |