DeMeVa at LeWiDi-2025: Modeling Perspectives with In-Context Learning and Label Distribution Learning
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866916946656624640 |
|---|---|
| author | Ignatev, Daniil Li, Nan Wong, Hugh Mee Dang, Anh Yaschuk, Shane Kaszefski |
| author_facet | Ignatev, Daniil Li, Nan Wong, Hugh Mee Dang, Anh Yaschuk, Shane Kaszefski |
| contents | This system paper presents the DeMeVa team's approaches to the third edition of the Learning with Disagreements shared task (LeWiDi 2025; Leonardelli et al., 2025). We explore two directions: in-context learning (ICL) with large language models, where we compare example sampling strategies; and label distribution learning (LDL) methods with RoBERTa (Liu et al., 2019b), where we evaluate several fine-tuning methods. Our contributions are twofold: (1) we show that ICL can effectively predict annotator-specific annotations (perspectivist annotations), and that aggregating these predictions into soft labels yields competitive performance; and (2) we argue that LDL methods are promising for soft label predictions and merit further exploration by the perspectivist community. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_09524 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | DeMeVa at LeWiDi-2025: Modeling Perspectives with In-Context Learning and Label Distribution Learning Ignatev, Daniil Li, Nan Wong, Hugh Mee Dang, Anh Yaschuk, Shane Kaszefski Computation and Language Machine Learning This system paper presents the DeMeVa team's approaches to the third edition of the Learning with Disagreements shared task (LeWiDi 2025; Leonardelli et al., 2025). We explore two directions: in-context learning (ICL) with large language models, where we compare example sampling strategies; and label distribution learning (LDL) methods with RoBERTa (Liu et al., 2019b), where we evaluate several fine-tuning methods. Our contributions are twofold: (1) we show that ICL can effectively predict annotator-specific annotations (perspectivist annotations), and that aggregating these predictions into soft labels yields competitive performance; and (2) we argue that LDL methods are promising for soft label predictions and merit further exploration by the perspectivist community. |
| title | DeMeVa at LeWiDi-2025: Modeling Perspectives with In-Context Learning and Label Distribution Learning |
| topic | Computation and Language Machine Learning |
| url | https://arxiv.org/abs/2509.09524 |