# Bidirectional Reactive Programming for Machine Learning

arXiv: https://arxiv.org/abs/2311.16977 · HTML: https://mlsanity.com/paper/2311.16977/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 10 of its 42 references that are in the map.
- First posted on arXiv 28 Nov 2023.

## Builds on (most cited first, 10 of 10)

- [Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift](https://mlsanity.com/paper/1502.03167/) (2015)
- [Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer](https://mlsanity.com/paper/1701.06538/) (2017)
- [Streaming keyword spotting on mobile devices](https://mlsanity.com/paper/2005.06720/) (2020)
- [Synthesizing Optimal Collective Algorithms](https://mlsanity.com/paper/2008.08708/) (2020)
- [Instead of Rewriting Foreign Code for Machine Learning, Automatically Synthesize Fast Gradients](https://mlsanity.com/paper/2010.01709/) (2020)
- [Trajectory Planning for Autonomous Vehicles Using Hierarchical Reinforcement Learning](https://mlsanity.com/paper/2011.04752/) (2020)
- [You Only Linearize Once: Tangents Transpose to Gradients](https://mlsanity.com/paper/2204.10923/) (2022)
- [Multirate Training of Neural Networks](https://mlsanity.com/paper/2106.10771/) (2021)
- [Deep learning models for predictive maintenance: a survey, comparison, challenges and prospect](https://mlsanity.com/paper/2010.03207/) (2020)
- [Extending TensorFlow's Semantics with Pipelined Execution](https://mlsanity.com/paper/1908.09291/) (2019)

## Abstract (arXiv)

Reactive languages are dedicated to the programming of systems which interact continuously and concurrently with their environment. Values take the form of unbounded streams modeling the (discrete) passing of time or the sequence of concurrent interactions. While conventional reactivity models recurrences forward in time, we introduce a symmetric reactive construct enabling backward recurrences. Constraints on the latter allow to make the implementation practical. Machine Learning (ML) systems provide numerous motivations for all of this: we demonstrate that reverse-mode automatic differentiation, backpropagation, batch normalization, bidirectional recurrent neural networks, training and reinforcement learning algorithms, are all naturally captured as bidirectional reactive programs.