用方向性类型提升语法泛化能力，显著改善位置变换任务表现。
On the Role of Directionality in Structural Generalization
- 采用带方向的CCG类型重构符号后端，支持确定性解析与线性解码
- 在SLOG测试中达75.9%精确匹配率，位置变换类任务提升29.9个百分点
- 方向性设计使性能瓶颈转向神经层，适配更强大编码器提升效果
多个SLOG测试类别明确涉及方向性差异（修饰语位置移动、论元提取位置），但此前最优模型AM-Parser使用的AM代数操作不包含方向信息。本文重新设计符号后端，基于带方向的CCG类型（确定性CKY + 单一线性解码器，30K可学习参数）。在相同BERT-base编码器下，系统取得75.9±6.4%的LF精确匹配率，优于AM-Parser（70.8±4.3%）。按SLOG自身类别分组分析，增益高度依赖方向性：CCG系统在全部5个位置移动类别上胜过AM-Parser（+29.9pp），而AM-Parser在全部6个递归深度类别上占优。将编码器替换为DeBERTa-v3-large后，准确率达到90.7±4.9%，其中递归深度类别的提升最为显著，与方向性带来的增益形成互补。方向性表示将性能瓶颈从符号层（AM-Parser在某些类别上0%上限）转移至神经层，后者可通过升级编码器持续优化。
原文摘要 · Abstract (English)
Several SLOG test categories explicitly involve directional distinctions (modifier position shifts, argument extraction positions), yet AM-Parser, the previous SOTA, uses an AM algebra whose operations do not encode direction. We redesign the symbolic backend around CCG directed types (deterministic CKY + single linear decoder, 30K learnable parameters). Under the same BERT-base encoder, the system achieves 75.9$\pm$6.4% LF exact match, surpassing AM-Parser (70.8$\pm$4.3%). Per SLOG's own category groupings, gains are highly directional: the CCG system outperforms AM-Parser on all 5 position-shift categories (+29.9pp), while AM-Parser outperforms on all 6 recursive-depth categories. Replacing the encoder with DeBERTa-v3-large yields 90.7$\pm$4.9%, with the largest encoder gains in recursive-depth categories, complementary to directionality's gains. Directional representations shift the bottleneck from the symbolic layer (AM-Parser's 0% category ceiling) to the neural layer, which improves with encoder upgrades.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。