Structure-Internalized Rule Language Model for Faithful Knowledge Graph Reasoning

Open paper
Benchmark Model Rank Results
link-prediction-on-wn18rrSIRLM (SFT)#11Hits@10: 0.648MRR: 0.556
link-prediction-on-wn18rrSIRLM (GRPO)#12Hits@10: 0.629MRR: 0.523
link-prediction-on-wn18rrSIRLM (PT)#43Hits@10: 0.556MRR: 0.461

Rank counts only results with a code link.