Understanding LLM Behavior in Multi-Target Cross-Lingual Summarization
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2026
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866916071911456768 |
|---|---|
| author | Ryu, Sangwon Liu, Yihong Wang, Mingyang Kim, Yunsu Ok, Jungseul Lee, Gary Geunbae Schuetze, Hinrich |
| author_facet | Ryu, Sangwon Liu, Yihong Wang, Mingyang Kim, Yunsu Ok, Jungseul Lee, Gary Geunbae Schuetze, Hinrich |
| contents | Multi-target cross-lingual text summarization (MTXLS), which summarizes a source document into multiple target languages, is increasingly important as users consume content in diverse languages, but remains underexplored. To address this gap, we introduce multi-target cross-lingual element-aware (MEA), a new MTXLS benchmark covering 24 target languages. We benchmark end-to-end and pipeline approaches across various LLMs and show that MTXLS performance still substantially lags behind English monolingual summarization. To better understand MTXLS in LLMs, we propose a layer-wise analysis framework for investigating how LLMs internally perform MTXLS. Our analyses suggest that translation and summarization behaviors emerge jointly within later layers rather than as distinctly decomposed stages. Most task-relevant processing occurs within these layers, and errors also tend to arise at similar depths. Motivated by these findings, we introduce an inference-time activation steering method that leverages hidden representations from English summarization to guide MTXLS generation. Experiments show that our method consistently improves MTXLS quality across target languages. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2606_01252 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Understanding LLM Behavior in Multi-Target Cross-Lingual Summarization Ryu, Sangwon Liu, Yihong Wang, Mingyang Kim, Yunsu Ok, Jungseul Lee, Gary Geunbae Schuetze, Hinrich Computation and Language Artificial Intelligence Multi-target cross-lingual text summarization (MTXLS), which summarizes a source document into multiple target languages, is increasingly important as users consume content in diverse languages, but remains underexplored. To address this gap, we introduce multi-target cross-lingual element-aware (MEA), a new MTXLS benchmark covering 24 target languages. We benchmark end-to-end and pipeline approaches across various LLMs and show that MTXLS performance still substantially lags behind English monolingual summarization. To better understand MTXLS in LLMs, we propose a layer-wise analysis framework for investigating how LLMs internally perform MTXLS. Our analyses suggest that translation and summarization behaviors emerge jointly within later layers rather than as distinctly decomposed stages. Most task-relevant processing occurs within these layers, and errors also tend to arise at similar depths. Motivated by these findings, we introduce an inference-time activation steering method that leverages hidden representations from English summarization to guide MTXLS generation. Experiments show that our method consistently improves MTXLS quality across target languages. |
| title | Understanding LLM Behavior in Multi-Target Cross-Lingual Summarization |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2606.01252 |