Back to Search
Start Over
On the Representational Capacity of Recurrent Neural Language Models
- Source :
- Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 7011-7034
- Publication Year :
- 2023
-
Abstract
- This work investigates the computational expressivity of language models (LMs) based on recurrent neural networks (RNNs). Siegelmann and Sontag (1992) famously showed that RNNs with rational weights and hidden states and unbounded computation time are Turing complete. However, LMs define weightings over strings in addition to just (unweighted) language membership and the analysis of the computational power of RNN LMs (RLMs) should reflect this. We extend the Turing completeness result to the probabilistic case, showing how a rationally weighted RLM with unbounded computation time can simulate any deterministic probabilistic Turing machine (PTM) with rationally weighted transitions. Since, in practice, RLMs work in real-time, processing a symbol at every time step, we treat the above result as an upper bound on the expressivity of RLMs. We also provide a lower bound by showing that under the restriction to real-time computation, such models can simulate deterministic real-time rational PTMs.<br />Comment: Added requirement for non-negative probabilities to definitions 2.3 and 3.1, fixed typos
- Subjects :
- Computer Science - Computation and Language
Computer Science - Machine Learning
Subjects
Details
- Database :
- arXiv
- Journal :
- Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 7011-7034
- Publication Type :
- Report
- Accession number :
- edsarx.2310.12942
- Document Type :
- Working Paper
- Full Text :
- https://doi.org/10.18653/v1/2023.emnlp-main.434