arXiv cs.CL· Sunjae Kwon, Suhaas Kiran DG, Zonghai Yao, Talha Chafekar, Deepali Singh, Hong Yu·· 12 小时前AI 评分27
SAGE：面向多模态语义对齐的释义增强推理评测基准
SAGE: Sense-Aligned Multimodal Grounding Evaluation with Gloss-Augmented Reasoning
AI 导读
研究者提出 SAGE，一个统一的 10 候选图像选择基准，将 VWSD、VerSe 和 WikiDiverse 重构为共享的以术语为中心的评测协议，覆盖名词、动词和实体三类语义对齐，并包含易混淆硬负样本与上下文替换的 VerSe 诊断。
正文 · 原文
Abstract:This paper introduces Sense-Aligned Grounding Evaluation (SAGE), a unified 10-candidate image-selection benchmark that reformulates VWSD, VerSe, and WikiDiverse into a shared term-centric protocol covering noun-, verb-, and entity-centered grounding. SAGE includes confusable hard negatives and a context-swapped VerSe diagnostic to probe target-term grounding beyond direct caption--image compatibility. We further propose Gloss-Augmented Reasoning (GAR), a training-free framework combining definition grounding with constraint-guided semantic verification. Experiments show that strong VWSD performance does not transfer uniformly across grounding settings. With Gemma3-12B, definition augmentation improves all three subsets, while constraint-guided reasoning improves VWSD and yields particularly strong gains on VerSe, but is less effective on WikiDiverse. Combining the two signals achieves the best macro-average performance, improving Hits@1 from 65.90 to 71.22. These results show that SAGE supports both unified evaluation and fine-grained analysis of explicit semantic reasoning across grounding settings.
| Subjects: | Computation and Language (cs.CL); Computer Vision and Pattern Recognition (cs.CV) |
| Cite as: | arXiv:2610.31838 [cs.CL] |
| (or arXiv:2610.31838v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2610.31838 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Sunjae Kwon [view email]
[v1]
Thu, 22 Oct 2026 19:23:18 UTC (23,922 KB)
来源：arXiv cs.CL · arxiv.org