Uncertainty-Aware Decoding with Minimum Bayes Risk
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866916646359138304 |
|---|---|
| author | Daheim, Nico Meister, Clara Möllenhoff, Thomas Gurevych, Iryna |
| author_facet | Daheim, Nico Meister, Clara Möllenhoff, Thomas Gurevych, Iryna |
| contents | Despite their outstanding performance in the majority of scenarios, contemporary language models still occasionally generate undesirable outputs, for example, hallucinated text. While such behaviors have previously been linked to uncertainty, there is a notable lack of methods that actively consider uncertainty during text generation. In this work, we show how Minimum Bayes Risk (MBR) decoding, which selects model generations according to an expected risk, can be generalized into a principled uncertainty-aware decoding method. In short, we account for model uncertainty during decoding by incorporating a posterior over model parameters into MBR's computation of expected risk. We show that this modified expected risk is useful for both choosing outputs and deciding when to abstain from generation and can provide improvements without incurring overhead. We benchmark different methods for learning posteriors and show that performance improves with prediction diversity. We release our code publicly. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_05318 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Uncertainty-Aware Decoding with Minimum Bayes Risk Daheim, Nico Meister, Clara Möllenhoff, Thomas Gurevych, Iryna Computation and Language Artificial Intelligence Machine Learning Despite their outstanding performance in the majority of scenarios, contemporary language models still occasionally generate undesirable outputs, for example, hallucinated text. While such behaviors have previously been linked to uncertainty, there is a notable lack of methods that actively consider uncertainty during text generation. In this work, we show how Minimum Bayes Risk (MBR) decoding, which selects model generations according to an expected risk, can be generalized into a principled uncertainty-aware decoding method. In short, we account for model uncertainty during decoding by incorporating a posterior over model parameters into MBR's computation of expected risk. We show that this modified expected risk is useful for both choosing outputs and deciding when to abstain from generation and can provide improvements without incurring overhead. We benchmark different methods for learning posteriors and show that performance improves with prediction diversity. We release our code publicly. |
| title | Uncertainty-Aware Decoding with Minimum Bayes Risk |
| topic | Computation and Language Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2503.05318 |