# Neural Search Space in Gboard Decoder

arXiv: https://arxiv.org/abs/2410.15575 · HTML: https://mlsanity.com/paper/2410.15575/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 8 of its 24 references that are in the map.
- First posted on arXiv 21 Oct 2024.

## Builds on (most cited first, 8 of 8)

- [Communication-Efficient Learning of Deep Networks from Decentralized Data](https://mlsanity.com/paper/1602.05629/) (2016)
- [Advances and Open Problems in Federated Learning](https://mlsanity.com/paper/1912.04977/) (2019)
- [Subword Regularization: Improving Neural Network Translation Models with Multiple Subword Candidates](https://mlsanity.com/paper/1804.10959/) (2018)
- [Federated Learning for Mobile Keyboard Prediction](https://mlsanity.com/paper/1811.03604/) (2018)
- [Federated Learning of N-gram Language Models](https://mlsanity.com/paper/1910.03432/) (2019)
- [Differentiable Weighted Finite-State Transducers](https://mlsanity.com/paper/2010.01003/) (2020)
- [Mobile Keyboard Input Decoding with Finite-State Transducers](https://mlsanity.com/paper/1704.03987/) (2017)
- [Approximating probabilistic models as weighted finite automata](https://mlsanity.com/paper/1905.08701/) (2019)

## Built on by (most cited first, 1 of 1)

- [FUTO Swipe: Layout-Agnostic Neural Swipe Decoding](https://mlsanity.com/paper/2606.25247/) (2026)

## Abstract (arXiv)

Gboard Decoder produces suggestions by looking for paths that best match input touch points on the context aware search space, which is backed by the language Finite State Transducers (FST). The language FST is currently an N-gram language model (LM). However, N-gram LMs, limited in context length, are known to have sparsity problem under device model size constraint. In this paper, we propose Neural Search Space which substitutes the N-gram LM with a Neural Network LM (NN-LM) and dynamically constructs the search space during decoding. Specifically, we integrate the long range context awareness of NN-LM into the search space by converting its outputs given context, into the language FST at runtime. This involves language FST structure redesign, pruning strategy tuning, and data structure optimizations. Online experiments demonstrate improved quality results, reducing Words Modified Ratio by [0.26%, 1.19%] on various locales with acceptable latency increases. This work opens new avenues for further improving keyboard decoding quality by enhancing neural LM more directly.