# SplaXBERT: Leveraging Mixed Precision Training and Context Splitting for Question Answering

arXiv: https://arxiv.org/abs/2412.05499 · HTML: https://mlsanity.com/paper/2412.05499/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 5 of its 9 references that are in the map.
- First posted on arXiv 7 Dec 2024.

## Builds on (most cited first, 5 of 5)

- [ALBERT: A Lite BERT for Self-supervised Learning of Language Representations](https://mlsanity.com/paper/1909.11942/) (2019)
- [Mixed Precision Training](https://mlsanity.com/paper/1710.03740/) (2017)
- [Highway Networks](https://mlsanity.com/paper/1505.00387/) (2015)
- [Ensemble ALBERT on SQuAD 2.0](https://mlsanity.com/paper/2110.09665/) (2021)
- [Comparative Study of Machine Learning Models and BERT on SQuAD](https://mlsanity.com/paper/2005.11313/) (2020)

## Abstract (arXiv)

SplaXBERT, built on ALBERT-xlarge with context-splitting and mixed precision training, achieves high efficiency in question-answering tasks on lengthy texts. Tested on SQuAD v1.1, it attains an Exact Match of 85.95% and an F1 Score of 92.97%, outperforming traditional BERT-based models in both accuracy and resource efficiency.