# Improving the Expressiveness of Deep Learning Frameworks with Recursion

arXiv: https://arxiv.org/abs/1809.00832 · HTML: https://mlsanity.com/paper/1809.00832/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 4 Sep 2018.

## Built on by (most cited first, 2 of 2)

- [SoD2: Statically Optimizing Dynamic Deep Neural Network](https://mlsanity.com/paper/2403.00176/) (2024)
- [Recursive Function Definitions in Static Dataflow Graphs and their Implementation in TensorFlow](https://mlsanity.com/paper/2410.20225/) (2024)

## Abstract (arXiv)

Recursive neural networks have widely been used by researchers to handle applications with recursively or hierarchically structured data. However, embedded control flow deep learning frameworks such as TensorFlow, Theano, Caffe2, and MXNet fail to efficiently represent and execute such neural networks, due to lack of support for recursion. In this paper, we add recursion to the programming model of existing frameworks by complementing their design with recursive execution of dataflow graphs as well as additional APIs for recursive definitions. Unlike iterative implementations, which can only understand the topological index of each node in recursive data structures, our recursive implementation is able to exploit the recursive relationships between nodes for efficient execution based on parallel computation. We present an implementation on TensorFlow and evaluation results with various recursive neural network models, showing that our recursive implementation not only conveys the recursive nature of recursive neural networks better than other implementations, but also uses given resources more effectively to reduce training and inference time.