用知识图谱生成更高质量的对比样本，提升大模型推理能力
KARMA: Knowledge graph-based Automated Reasoning Materialization and Alignment
- 基于知识图谱枚举路径并转化为对齐槽位的对比样本
- 在生物、计算机、化学多个基准上超越基线模型
- 适合需要高精度推理的领域应用
基于模板的对比合成方法虽可扩展，但候选样本仅在少数实体槽位上有差异，而序列级优化则将监督信号分散到大部分共享模板中。我们将其形式化为分辨率不匹配问题，并提出KARMA：在领域知识图谱上枚举符合模式约束的路径，并将其转写为槽位对齐的对比候选。槽位并行对齐（SPA）采用解耦的槽位级目标，将偏好监督引导至具有区分性的实体槽位，槽位感知的掩码注意力可作为可选的打包评估实现方式。在生物医学、计算机科学和化学多个基准上，KARMA优于基础大模型和同数据微调基线，在序列与标记级偏好方法中表现也具竞争力。
原文摘要 · Abstract (English)
Template-based contrastive synthesis is scalable, but its candidates often differ only in a few entity-slots while sequence-level optimization spreads supervision over mostly shared templates. We formalize this as the Resolution Mismatch Problem and propose KARMA, which enumerates schema-constrained paths over domain knowledge graphs and verbalizes them into slot-aligned contrastive candidates. Slot-Parallel Alignment (SPA) then applies a decoupled slot-level objective to route preference supervision to discriminative entity-slots, with slot-aware masked attention serving as an optional packed-evaluation implementation. Across biomedical, computer-science, and chemistry benchmarks, KARMA outperforms base LLM and same-data SFT baselines, and compares favorably with sequence and token-level preference methods.
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。