Learning to reason is hard
Samy Bengio
Senior Director, AI and Machine Learning Research, Apple
ISC Summer School on Large Language Models: Science and Stakes June 3-14, 2024
Tues, June 11, 9am-10:30am EDT
Abstract: Reasoning is the action of drawing conclusions efficiently by composing learned concepts. In this presentation I’ll give a few examples illustrating why it is hard to learn to reason with current machine learning approaches. I will describe a general framework (generalization of the unseen) that characterizes most reasoning problems and out-of-distribution generalization in general, and give insights about intrinsic biases of current models. I will then present the specific problem of length generalization and why some instances can be solved by models like Transformers and some cannot.
SAMY BENGIO is a senior director of machine learning research at Apple since 2021. His research interests span areas of machine learning such as deep architectures, representation learning, vision and language processing and more recently, reasoning. He co-wrote the well-known open-source Torch machine learning library.
Boix-Adsera, E., Saremi, O., Abbe, E., Bengio, S., Littwin, E., & Susskind, J. (2023). When can transformers reason with abstract symbols? arXiv preprint arXiv:2310.09753.ICLR 2024
Zhou,, E., Razin, N., Saremi, O., Susskind, J., … & Nakkiran, P. (2023). What algorithms can transformers learn? a study in length generalization. arXiv preprint arXiv:2310.16028. ICLR 2024
Abbe, E., Bengio, S., Lotfi, A., & Rizk, K. (2023, July). Generalization on the unseen, logic reasoning and degree curriculum. In International Conference on Machine Learning (pp. 31-60). PMLR. ICML 2023.
28 сен 2024