ALIGNed-LLM aligns knowledge graph entity embeddings with language model text embeddings through a trainable projection layer, improving question answering accuracy on KG-derived datasets.
URL: https://www.anthropic.com/product
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Aligning Knowledge Graphs and Language Models for Factual Accuracy
ALIGNed-LLM aligns knowledge graph entity embeddings with language model text embeddings through a trainable projection layer, improving question answering accuracy on KG-derived datasets.