{"type":"ScholarlyArticle","title":"Causal Graph Modeling with Deep Neural Engines for Strong Abstract Reasoning in Language and Vision","authors":[{"name":"Gaël Gendron","slug":"gael-gendron","given":"Gaël","family":"Gendron","affiliation":"University of Auckland"}],"abstract":"Deep learning (DL) relies on discovering correlation patterns in low-level data and aggregating the information to solve a task. Despite success in a wide variety of applications, ranging from natural language to vision tasks, the learned patterns are often brittle and do not transfer out of the training data distribution (i.e. to different domains). Causality theory proposes methods to discover and estimate cause-effect relationships beyond correlations. Its powerful inference frameworks have been recently highlighted as a potential way to improve the lack of out-of-distribution generalisation in deep neural networks. However, their applications to deep learning problems remain largely under-explored. Our work attempts to bridge this gap and apply causal graphical models to abstract and causal reasoning problems in natural language and vision, requiring strong generalisation abilities beyond correlations. We integrate causal graph modelling methods into deep vision networks and Large Language Models to improve their capacity to perform strong and out-of-distribution reasoning on complex abstract problems.","doi":"10.24963/ijcai.2024/960","url":"https://ijcai.org/proceedings/2024/960","pdf":"https://ijcai.org/proceedings/2024/0960.pdf","markdown":"https://ijcai.org/proceedings/2024/960.md","bibtex":"https://ijcai.org/proceedings/2024/bibtex/960","volume":{"slug":"2024","title":"Proceedings of the Thirty-Third International Joint Conference on Artificial Intelligence","short_title":"IJCAI 2024","url":"https://ijcai.org/proceedings/2024/","year":2024,"location":"Jeju, Korea","dates":"3-9 August 2024","editors":"Kate Larson","isbn":"978-1-956792-04-1","publisher":"International Joint Conferences on Artificial Intelligence Organization","doi_prefix":"10.24963/ijcai.2024"},"section":"Doctoral Consortium","track":"Doctoral Consortium","number":960,"pages":{"first":8490,"last":8491},"keywords":["DC: Machine Learning","DC: Computer Vision","DC: Natural Language Processing","DC: Knowledge Representation and Reasoning"],"videos":[],"references":[{"position":1,"raw":"[Gendron et al., 2023a] Gaël Gendron, Qiming Bao, Michael Witbrock, and Gillian Dobbie. Large language models are not abstract reasoners. CoRR, abs/2305.19555, 2023.","title":"Large language models are not abstract reasoners","authors":"Gendron","year":"2023","venue":null,"doi":null,"url":null,"internal_url":null},{"position":2,"raw":"[Gendron et al., 2023b] Gaël Gendron, Yang Chen, Mitchell Rogers, Yiping Liu, Mihailo Azhar, Shahrokh Heidari, David Arturo Soriano Valdez, Kobe Knowles, Padriac O'Leary, Simon Eyre, et al. Behaviour modelling of so- cial animals via causal structure discovery and graph neu- ral networks. arXiv preprint arXiv:2312.14333, 2023.","title":"Behaviour modelling of social animals via causal structure discovery and graph neural networks","authors":"Gendron","year":"2023","venue":null,"doi":null,"url":null,"internal_url":null},{"position":3,"raw":"[Gendron et al., 2023c] Gaël Gendron, Michael Witbrock, and Gillian Dobbie. Disentanglement of latent represen- tations via causal interventions. In Proceedings of the Thirty-Second International Joint Conference on Artificial Intelligence, IJCAI 2023, 2023.","title":"Disentanglement of latent representations via causal interventions","authors":"Gendron","year":"2023","venue":"Proceedings of the Thirty-Second International Joint Conference on Artificial Intelligence, IJCAI 2023","doi":null,"url":null,"internal_url":"/proceedings/2023/361"},{"position":4,"raw":"[Gendron et al., 2024] Gaël Gendron, Bao Trung Nguyen, Alex Yuxuan Peng, Michael Witbrock, and Gillian Dob- bie. Can large language models learn independent causal mechanisms? CoRR, abs/2402.02636, 2024.","title":"Can large language models learn independent causal mechanisms?","authors":"Gaël Gendron, Bao Trung Nguyen, Alex Yuxuan Peng, Michael Witbrock, Gillian Dobbie","year":"2024","venue":null,"doi":null,"url":null,"internal_url":null},{"position":5,"raw":"[Goyal and Bengio, 2022] Anirudh Goyal and Yoshua Ben- gio. Inductive biases for deep learning of higher- level cognition. Proceedings of the Royal Society A, 478(2266):20210068, 2022.","title":"Inductive biases for deep learning of higherlevel cognition","authors":"Anirudh Goyal, Yoshua Bengio","year":"2022","venue":"Proceedings of the Royal Society A","doi":null,"url":null,"internal_url":null},{"position":6,"raw":"[Li and Flanigan, 2024] Changmao Li and Jeffrey Flanigan. Task contamination: Language models may not be few- shot anymore. In Thirty-Eighth AAAI Conference on Arti- ficial Intelligence, AAAI, 2024.","title":"Task contamination: Language models may not be fewshot anymore","authors":"Li, Flanigan, Changmao Li, Jeffrey Flanigan","year":"2024","venue":"Thirty-Eighth AAAI Conference on Artificial Intelligence","doi":null,"url":null,"internal_url":null},{"position":7,"raw":"[Pearl, 2009] Judea Pearl. Causality. Cambridge university press, 2009.","title":"Judea Pearl. Causality","authors":"Pearl","year":"2009","venue":null,"doi":null,"url":null,"internal_url":null},{"position":8,"raw":"[Schölkopf et al., 2021] Bernhard Schölkopf, Francesco Lo- catello, Stefan Bauer, Nan Rosemary Ke, Nal Kalchbren- ner, Anirudh Goyal, and Yoshua Bengio. Toward causal representation learning. Proc. IEEE, 109(5):612-634, 2021.","title":"Toward causal representation learning","authors":"Bernhard Schölkopf, Francesco Locatello, Stefan Bauer, Nan Rosemary Ke, Nal Kalchbrenner, Anirudh Goyal, Yoshua Bengio","year":"2021","venue":"Proc. IEEE","doi":null,"url":null,"internal_url":null}],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}