nchellwig at SemEval-2026 Task 3: Self-Consistent Structured Generation (SCSG) for Dimensional Aspect-Based Sentiment Analysis using Large Language Models
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910037897641984 |
|---|---|
| author | Hellwig, Nils Constantin Fehle, Jakob Kruschwitz, Udo Wolff, Christian |
| author_facet | Hellwig, Nils Constantin Fehle, Jakob Kruschwitz, Udo Wolff, Christian |
| contents | We present Self-Consistent Structured Generation (SCSG) for Dimensional Aspect-Based Sentiment Analysis in SemEval-2026 Task 3 (Track A). SCSG enhances prediction reliability by executing a LoRA-adapted large language model multiple times per instance, retaining only tuples that achieve a majority consensus across runs. To mitigate the computational overhead of multiple forward passes, we leverage vLLM's PagedAttention mechanism for efficient key--value cache reuse. Evaluation across 6 languages and 8 language--domain combinations demonstrates that self-consistency with 15 executions yields statistically significant improvements over single-inference prompting, with our system (leveraging Gemma 3) ranking in the top seven across all settings, achieving second place on three out of four English subsets and first place on Tatar-Restaurant for DimASTE. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_01788 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | nchellwig at SemEval-2026 Task 3: Self-Consistent Structured Generation (SCSG) for Dimensional Aspect-Based Sentiment Analysis using Large Language Models Hellwig, Nils Constantin Fehle, Jakob Kruschwitz, Udo Wolff, Christian Computation and Language We present Self-Consistent Structured Generation (SCSG) for Dimensional Aspect-Based Sentiment Analysis in SemEval-2026 Task 3 (Track A). SCSG enhances prediction reliability by executing a LoRA-adapted large language model multiple times per instance, retaining only tuples that achieve a majority consensus across runs. To mitigate the computational overhead of multiple forward passes, we leverage vLLM's PagedAttention mechanism for efficient key--value cache reuse. Evaluation across 6 languages and 8 language--domain combinations demonstrates that self-consistency with 15 executions yields statistically significant improvements over single-inference prompting, with our system (leveraging Gemma 3) ranking in the top seven across all settings, achieving second place on three out of four English subsets and first place on Tatar-Restaurant for DimASTE. |
| title | nchellwig at SemEval-2026 Task 3: Self-Consistent Structured Generation (SCSG) for Dimensional Aspect-Based Sentiment Analysis using Large Language Models |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2603.01788 |