# Later-stage Minimum Bayes-Risk Decoding for Neural Machine Translation

arXiv: https://arxiv.org/abs/1704.03169 · HTML: https://mlsanity.com/paper/1704.03169/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 11 Apr 2017.

## Built on by (most cited first, 8 of 8)

- [Sampling-Based Approximations to Minimum Bayes Risk Decoding for Neural Machine Translation](https://mlsanity.com/paper/2108.04718/) (2021)
- [Understanding the Properties of Minimum Bayes Risk Decoding in Neural Machine Translation](https://mlsanity.com/paper/2105.08504/) (2021)
- [Quality-Aware Decoding for Neural Machine Translation](https://mlsanity.com/paper/2205.00978/) (2022)
- [Transcormer: Transformer for Sentence Scoring with Sliding Language Modeling](https://mlsanity.com/paper/2205.12986/) (2022)
- [DC-MBR: Distributional Cooling for Minimum Bayesian Risk Decoding](https://mlsanity.com/paper/2212.04205/) (2022)
- [RMBR: A Regularized Minimum Bayes Risk Reranking Framework for Machine Translation](https://mlsanity.com/paper/2203.00201/) (2022)
- [On the True Distribution Approximation of Minimum Bayes-Risk Decoding](https://mlsanity.com/paper/2404.00752/) (2024)
- [Unveiling the Power of Source: Source-based Minimum Bayes Risk Decoding for Neural Machine Translation](https://mlsanity.com/paper/2406.11632/) (2024)

## Abstract (arXiv)

For extended periods of time, sequence generation models rely on beam search algorithm to generate output sequence. However, the correctness of beam search degrades when the a model is over-confident about a suboptimal prediction. In this paper, we propose to perform minimum Bayes-risk (MBR) decoding for some extra steps at a later stage. In order to speed up MBR decoding, we compute the Bayes risks on GPU in batch mode. In our experiments, we found that MBR reranking works with a large beam size. Later-stage MBR decoding is shown to outperform simple MBR reranking in machine translation tasks.