GLaM: Fine-Tuning Large Language Models for Domain Knowledge Graph Alignment via Neighborhood Partitioning and Generative Subgraph Encoding (2402.06764v3)
Abstract: Integrating LLMs with knowledge graphs derived from domain-specific data represents an important advancement towards more powerful and factual reasoning. As these models grow more capable, it is crucial to enable them to perform multi-step inferences over real-world knowledge graphs while minimizing hallucination. While LLMs excel at conversation and text generation, their ability to reason over domain-specialized graphs of interconnected entities remains limited. For example, can we query a LLM to identify the optimal contact in a professional network for a specific goal, based on relationships and attributes in a private database? The answer is no--such capabilities lie beyond current methods. However, this question underscores a critical technical gap that must be addressed. Many high-value applications in areas such as science, security, and e-commerce rely on proprietary knowledge graphs encoding unique structures, relationships, and logical constraints. We introduce a fine-tuning framework for developing Graph-aligned LLMs (GLaM) that transforms a knowledge graph into an alternate text representation with labeled question-answer pairs. We demonstrate that grounding the models in specific graph-based knowledge expands the models' capacity for structure-based reasoning. Our methodology leverages the large-LLM's generative capabilities to create the dataset and proposes an efficient alternate to retrieval-augmented generation styled methods.
- The Reversal Curse: LLMs trained on” A is B” fail to learn” B is A”. arXiv preprint arXiv:2309.12288.
- Bodenreider, O. 2004. The unified medical language system (UMLS): integrating biomedical terminology. Nucleic acids research, 32(suppl_1): D267–D270.
- Graph-based Multilingual Language Model: Leveraging Product Relations for Search Relevance. In Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, 2789–2799.
- Talk like a graph: Encoding graphs for large language models. arXiv preprint arXiv:2310.04560.
- Reasoning with language model is planning with world model. arXiv preprint arXiv:2305.14992.
- Deberta: Decoding-enhanced bert with disentangled attention. arXiv preprint arXiv:2006.03654.
- Structgpt: A general framework for large language model to reason over structured data. arXiv preprint arXiv:2305.09645.
- Large Language Models on Graphs: A Comprehensive Survey. arXiv preprint arXiv:2312.02783.
- Retrieval-augmented generation for knowledge-intensive nlp tasks. Advances in Neural Information Processing Systems, 33: 9459–9474.
- Towards graph foundation models: A survey and beyond. arXiv preprint arXiv:2310.11829.
- Capabilities of gpt-4 on medical challenge problems. arXiv preprint arXiv:2303.13375.
- Training language models to follow instructions with human feedback. Advances in Neural Information Processing Systems, 35: 27730–27744.
- Unifying Large Language Models and Knowledge Graphs: A Roadmap. arXiv preprint arXiv:2306.08302.
- Deepspeed: System optimizations enable training deep learning models with over 100 billion parameters. In Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, 3505–3506.
- Improving multi-hop question answering over knowledge graphs using knowledge base embeddings. In Proceedings of the 58th annual meeting of the association for computational linguistics, 4498–4507.
- Outrageously large neural networks: The sparsely-gated mixture-of-experts layer. arXiv preprint arXiv:1701.06538.
- Monte Carlo Thought Search: Large Language Model Querying for Complex Scientific Reasoning in Catalyst Design. arXiv preprint arXiv:2310.14420.
- Think-on-graph: Deep and responsible reasoning of large language model with knowledge graph. arXiv preprint arXiv:2307.07697.
- ArnetMiner: Extraction and Mining of Academic Social Networks. In KDD’08, 990–998.
- Graph neural prompting with large language models. arXiv preprint arXiv:2309.15427.
- Llama 2: Open foundation and fine-tuned chat models. arXiv preprint arXiv:2307.09288.
- Finetuned language models are zero-shot learners. arXiv preprint arXiv:2109.01652.
- Tree of thoughts: Deliberate problem solving with large language models. arXiv preprint arXiv:2305.10601.
- Deep bidirectional language-knowledge graph pretraining. Advances in Neural Information Processing Systems, 35: 37309–37323.
- Bertscore: Evaluating text generation with bert. arXiv preprint arXiv:1904.09675.
- Mixture-of-experts with expert choice routing. Advances in Neural Information Processing Systems, 35: 7103–7114.
Collections
Sign up for free to add this paper to one or more collections.
Paper Prompts
Sign up for free to create and run prompts on this paper using GPT-5.