发表机构
Imperial College London（帝国理工学院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本文针对立场感知的论证检索任务，发现现有稠密嵌入模型存在主题偏好、过度关注极性关键词等问题，通过引入诊断指标并提出结合平衡论证课程与立场反转论证的数据中心方案，缓解了模型过度纠正问题，提升了立场感知的论证检索效果。
AI 中文摘要
在计算论证领域，获取明确支持或攻击给定主张的论证是下游推理任务的关键前提。当使用语义搜索方法检索这些支持性与攻击型论证时，需同时评估其与目标主张的主题相关性，以及对主张的（正向或负向）立场正确性。本文探讨驱动现代检索流水线的稠密嵌入模型（以下简称模型）如何成为整合这种双重评估的语义搜索基础。实验显示，现有模型在非对称推理中表现不佳，存在强烈的主题重叠偏好，同时忽略指令性立场；还发现通过对比训练纠正该偏差会触发新的失效模式，即模型过度纠正，过度关注极性关键词（如“支持”或“反驳”）而牺牲语义主题。因此，本文引入诊断性词消融指标量化该现象，并提出以数据为中心的解决方案：通过实施平衡的论证课程，结合大语言模型（LLM）增强的立场反转论证，迫使嵌入模型学习更深入的方向逻辑，而非利用表面词汇捷径。评估表明，对于足够强大的模型，该方法可缓解观察到的过度纠正问题，在立场感知的论证检索中实现进一步提升。
英文摘要
In computational argumentation, obtaining arguments that explicitly support or attack given claims is a critical precursor to downstream reasoning tasks. When these supporting and attacking arguments are to be retrieved using semantic search methods, they need to be assessed for topic-relevance to the claims of interest as well as for correctness of their (positive or negative) stance towards the claims. In this paper we explore how dense embedding models (hereafter, models), powering modern retrieval pipelines, can serve as the basis of semantic search incorporating this dual assessment. We show experimentally that existing models struggle with asymmetric reasoning, exhibiting a strong bias toward topical overlap while ignoring instructional stance. We also show that correcting this bias via contrastive training triggers a new failure mode where models over-correct, over-fixating on polarity keywords (e.g., "supports" or "refutes") at the expense of the semantic topic. We thus introduce diagnostic word-ablation metrics to quantify this phenomenon and propose a data-centric solution. By implementing a balanced argument curriculum alongside LLM-augmented, stance-inverted arguments, we force the (embedding) models to learn deeper directional logic rather than exploiting superficial lexical shortcuts. Our evaluation demonstrates that, for sufficiently powerful models, this approach can alleviate the observed overcorrection, achieving further improvements in stance-aware argument retrieval.
CommentsCMNA'26
