Steering Large Language Models for Machine Translation Personalization
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2025
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866912645363269632 |
|---|---|
| author | Scalena, Daniel Sarti, Gabriele Bisazza, Arianna Fersini, Elisabetta Nissim, Malvina |
| author_facet | Scalena, Daniel Sarti, Gabriele Bisazza, Arianna Fersini, Elisabetta Nissim, Malvina |
| contents | Large language models have simplified the production of personalized translations reflecting predefined stylistic constraints. However, these systems still struggle when stylistic requirements are implicitly represented by a set of examples, such as texts produced by a specific human translator. In this work, we explore various strategies for personalizing automatically generated translations when few examples are available, with a focus on the challenging domain of literary translation. We begin by determining the feasibility of the task and how style information is encoded within model representations. Then, we evaluate various prompting strategies and inference-time interventions for steering model generations towards a personalized style, with a particular focus on contrastive steering with sparse autoencoder (SAE) latents to identify salient personalization properties. We demonstrate that contrastive SAE steering yields robust style conditioning and translation quality, resulting in higher inference-time computational efficiency than prompting approaches. We further examine the impact of steering on model activations, finding that layers encoding personalization properties are impacted similarly by prompting and SAE steering, suggesting a similar mechanism at play. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2505_16612 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Steering Large Language Models for Machine Translation Personalization Scalena, Daniel Sarti, Gabriele Bisazza, Arianna Fersini, Elisabetta Nissim, Malvina Computation and Language Artificial Intelligence Machine Learning Large language models have simplified the production of personalized translations reflecting predefined stylistic constraints. However, these systems still struggle when stylistic requirements are implicitly represented by a set of examples, such as texts produced by a specific human translator. In this work, we explore various strategies for personalizing automatically generated translations when few examples are available, with a focus on the challenging domain of literary translation. We begin by determining the feasibility of the task and how style information is encoded within model representations. Then, we evaluate various prompting strategies and inference-time interventions for steering model generations towards a personalized style, with a particular focus on contrastive steering with sparse autoencoder (SAE) latents to identify salient personalization properties. We demonstrate that contrastive SAE steering yields robust style conditioning and translation quality, resulting in higher inference-time computational efficiency than prompting approaches. We further examine the impact of steering on model activations, finding that layers encoding personalization properties are impacted similarly by prompting and SAE steering, suggesting a similar mechanism at play. |
| title | Steering Large Language Models for Machine Translation Personalization |
| topic | Computation and Language Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2505.16612 |