Transformers Implement First-Order Logic with Majority QuantifiersDownload PDF

Published: 01 Feb 2023, Last Modified: 13 Feb 2023Submitted to ICLR 2023Readers: Everyone
Keywords: transformers, complexity theory, logic, interpretability
TL;DR: Transformers can be translated to formulae in first-order logic with majority quantifiers that compute the same function.
Abstract: Characterizing the implicit structure of the computation within neural networks is a foundational problem in the area of deep learning interpretability. Can their inner decision process be captured symbolically in some familiar logic? We show that any transformer neural network can be translated into an equivalent fixed-size first-order logic formula which may also use majority quantifiers. The idea is to simulate transformers with highly uniform threshold circuits and leverage known theoretical connections between circuits and logic. Our findings also reveal the surprising fact that the entire transformer computation can be reduced merely to the division of two (large) integers. While our results are most pertinent for transformers, they apply equally to a broader class of neural network architectures, namely those with a fixed-depth uniform computation graph made up of standard neural net components, which includes feedforward and convolutional networks.
Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.
No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.
Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics
Submission Guidelines: Yes
Please Choose The Closest Area That Your Submission Falls Into: Theory (eg, control theory, learning theory, algorithmic game theory)
12 Replies

Loading