Potential methods for quantifying alignment between LLMs and KGs include contrastive probing with synthetic counterfactuals or topology-aware alignment losses.

Authors

• Person: Not available Organization: arXiv
  Large Language Models Meet Knowledge Graphs for Question ...

Sources

• Large Language Models Meet Knowledge Graphs for Question ... arxiv.org arXiv via serper

Referenced by nodes (2)

• Large Language Models concept
• KGs concept