Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1]
Minimal Width for Universal Property of Deep RNN
Authors: Chang hoon Song, Geonho Hwang, Jun ho Lee, Myungjoo Kang
JMLR 2023 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Theoretical | In this study, we prove the universality of deep narrow RNNs and show that the upper bound of the minimum width for universality can be independent of the length of the data. Specifically, we show a deep RNN with Re LU activation can approximate any continuous function or Lp function with the widths dx + dy + 3 and max{dx + 1, dy}, respectively... In addition, we prove the universality of other recurrent networks, such as bidirectional RNNs. |
| Researcher Affiliation | Academia | Chang hoon Song EMAIL Department of Mathematical Science Seoul National University, Geonho Hwang EMAIL Department of Mathematical Science Seoul National University, Jun ho Lee EMAIL Kongju National Universality, Myungjoo Kang EMAIL Department of Mathematical Science Seoul National University |
| Pseudocode | No | The paper contains mathematical definitions, theorems, lemmas, and proofs. There are no sections explicitly labeled 'Pseudocode' or 'Algorithm', nor are there any structured algorithm blocks presenting procedural steps in a code-like format. |
| Open Source Code | No | The paper does not contain any statements about releasing code, links to code repositories, or mentions of supplementary materials with code for the methodology described. |
| Open Datasets | No | This paper is theoretical in nature, focusing on proving universal approximation theorems for RNNs. It does not describe or utilize any datasets for experimental evaluation. |
| Dataset Splits | No | The paper is theoretical and does not conduct experiments using datasets, therefore, no dataset split information is provided. |
| Hardware Specification | No | This paper is purely theoretical, focusing on mathematical proofs and analysis. It does not describe any experimental setup or the hardware used for running experiments. |
| Software Dependencies | No | The paper is theoretical and does not involve experimental implementation. Therefore, it does not specify any software dependencies or their version numbers. |
| Experiment Setup | No | The paper is theoretical, presenting universal approximation theorems and proofs rather than empirical studies. Consequently, it does not include details regarding experimental setup, hyperparameters, or training configurations. |