Knowledge graph reasoning (KGR) infers missing facts, with recent advances increasingly harnessing the semantic priors and reasoning abilities of Large Language Models (LLMs). However, prevailing generative paradigms are prone to memorizing surface-level co-occurrences rather than learning genuine relational semantics, limiting out-of-distribution generalization. To address this, we propose RADAR, which reformulates KGR from generative pattern matching to discriminative relational reasoning. We recast KGR as discriminative entity selection, where reinforcement learning enforces relative entity separability beyond token-likelihood imitation. Leveraging this separability, inference operates directly in representation space, ensuring consistency with the discriminative optimization and bypassing generation-induced hallucinations. Across four benchmarks, RADAR achieves 5-6% relative gains on link prediction and triple classification over strong LLM baselines, while increasing task-relevant mutual information in intermediate representations by 62.9%, indicating more robust and transferable relational reasoning.

翻译：知识图谱推理旨在推断缺失事实，近期研究日益利用大语言模型的语义先验与推理能力。然而，当前主流生成式范式倾向于记忆表层共现模式而非学习真实关系语义，限制了分布外泛化性能。为此，我们提出RADAR方法，将知识图谱推理从生成式模式匹配重构为判别式关系推理。我们将知识图谱推理重新定义为判别式实体选择任务，通过强化学习增强实体间相对可分离性，超越基于词元似然的模仿学习。利用这种可分离性，推理过程直接在表征空间中进行，确保与判别式优化目标的一致性，并规避生成过程引发的幻觉问题。在四个基准测试中，RADAR在链接预测和三元组分类任务上相较强基线大语言模型实现5-6%的相对性能提升，同时使中间表征中任务相关的互信息量增加62.9%，表明其具备更鲁棒且可迁移的关系推理能力。