Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
41 tokens/sec
GPT-4o
59 tokens/sec
Gemini 2.5 Pro Pro
41 tokens/sec
o3 Pro
7 tokens/sec
GPT-4.1 Pro
50 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

GraphWiz: An Instruction-Following Language Model for Graph Problems (2402.16029v5)

Published 25 Feb 2024 in cs.CL

Abstract: LLMs have achieved impressive success across several fields, but their proficiency in understanding and resolving complex graph problems is less explored. To bridge this gap, we introduce GraphInstruct, a novel and comprehensive instruction-tuning dataset designed to equip LLMs with the ability to tackle a broad spectrum of graph problems using explicit reasoning paths. Utilizing GraphInstruct, we build GraphWiz, an open-source LLM capable of resolving various graph problem types while generating clear reasoning processes. To enhance the model's capability and reliability, we incorporate the Direct Preference Optimization (DPO) framework into the graph problem-solving context. The enhanced model, GraphWiz-DPO, achieves an average accuracy of 65% across nine tasks with different complexity levels, surpassing GPT-4 which has an average accuracy of 43.8%. Moreover, our research delves into the delicate balance between training data volume and model performance, highlighting the potential for overfitting with increased data. We also explore the transferability of the model's reasoning ability across different graph tasks, indicating the model's adaptability and practical application potential. Our investigation offers a new blueprint and valuable insights for developing LLMs specialized in graph reasoning and problem-solving.

PDF HTML Abstract
Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize File Document Download Save Streamline Icon: https://streamlinehq.com PDF File Document Download Save Streamline Icon: https://streamlinehq.com Markdown Bookmark Chat Bubble Oval Streamline Icon: https://streamlinehq.com Chat (Pro)
Definition Search Book Streamline Icon: https://streamlinehq.com
References (49)
  1. Palm 2 technical report. arXiv preprint arXiv:2305.10403, 2023.
  2. Document clustering: Tf-idf approach. In 2016 International Conference on Electrical, Electronics, and Optimization Techniques (ICEEOT), pages 61–66, 2016. doi: 10.1109/ICEEOT.2016.7754750.
  3. Graphllm: Boosting graph reasoning ability of large language model. arXiv preprint arXiv:2310.05845, 2023.
  4. Large language models meet harry potter: A dataset for aligning dialogue agents with characters. In Findings of the Association for Computational Linguistics: EMNLP 2023, pages 8506–8520, Singapore, December 2023a.
  5. Beyond surface: Probing llama across scales and layers. arXiv preprint arXiv:2312.04333, 2023b.
  6. Breaking language barriers in multilingual mathematical reasoning: Insights and observations. arXiv preprint arXiv:2310.20246, 2023c.
  7. Exploring the potential of large language models (llms) in learning on graphs. arXiv preprint arXiv:2307.03393, 2023d.
  8. Label-free node classification on graphs with large language models (llms). arXiv preprint arXiv:2310.04668, 2023e.
  9. Training verifiers to solve math word problems. CoRR, abs/2110.14168, 2021.
  10. On the evolution of random graphs. Publ. math. inst. hung. acad. sci, 5(1):17–60, 1960.
  11. Talk like a graph: Encoding graphs for large language models. In NeurIPS 2023 Workshop: New Frontiers in Graph Learning, 2023.
  12. Pal: Program-aided language models. In International Conference on Machine Learning, pages 10764–10799. PMLR, 2023.
  13. Exploring network structure, dynamics, and function using networkx. Technical report, Los Alamos National Lab.(LANL), Los Alamos, NM (United States), 2008.
  14. Harnessing explanations: Llm-to-lm interpreter for enhanced text-attributed graph representation learning. arXiv preprint arXiv:2305.19523, 2023.
  15. Can llms effectively leverage graph structural information: When and why. arXiv preprint arXiv:2309.16595, 2023.
  16. Survey of hallucination in natural language generation. ACM Computing Surveys, 2023. URL https://dl.acm.org/doi/10.1145/3571730.
  17. Mistral 7b. arXiv preprint arXiv:2310.06825, 2023.
  18. Mixtral of experts. arXiv preprint arXiv:2401.04088, 2024.
  19. The impact of reasoning step length on large language models. arXiv preprint arXiv:2401.04925, 2024.
  20. Segment anything. arXiv preprint arXiv:2304.02643, 2023.
  21. Review on determining number of cluster in k-means clustering. International Journal, 1(6):90–95, 2013.
  22. Retrieval-augmented generation for knowledge-intensive nlp tasks. 2020.
  23. A survey of graph meets large language model: Progress and future directions. arXiv preprint arXiv:2311.12399, 2023.
  24. Visual instruction tuning. arXiv preprint arXiv:2304.08485, 2023.
  25. OpenAI. GPT-4 technical report. Arxiv, 2023. URL https://doi.org/10.48550/arXiv.2303.08774.
  26. Training language models to follow instructions with human feedback. Advances in Neural Information Processing Systems, 35:27730–27744, 2022.
  27. Direct preference optimization: Your language model is secretly a reward model. arXiv preprint arXiv:2305.18290, 2023.
  28. Sentence-BERT: Sentence embeddings using Siamese BERT-networks. In Kentaro Inui, Jing Jiang, Vincent Ng, and Xiaojun Wan, editors, Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP), pages 3982–3992, Hong Kong, China, November 2019. Association for Computational Linguistics. doi: 10.18653/v1/D19-1410. URL https://aclanthology.org/D19-1410.
  29. Bloom: A 176b-parameter open-access multilingual language model. arXiv preprint arXiv:2211.05100, 2022.
  30. Proximal policy optimization algorithms. arXiv preprint arXiv:1707.06347, 2017.
  31. Graphgpt: Graph instruction tuning for large language models. arXiv preprint arXiv:2310.13023, 2023.
  32. Stanford alpaca: An instruction-following llama model. https://github.com/tatsu-lab/stanford_alpaca, 2023.
  33. Llama: Open and efficient foundation language models. arXiv preprint arXiv:2302.13971, 2023a.
  34. Llama 2: Open foundation and fine-tuned chat models. arXiv preprint arXiv:2307.09288, 2023b.
  35. Solving olympiad geometry without human demonstrations. Nature, 625(7995):476–482, 2024.
  36. Can language models solve graph problems in natural language? In Thirty-seventh Conference on Neural Information Processing Systems, 2023. URL https://openreview.net/forum?id=UDqHhbqYJV.
  37. Self-consistency improves chain of thought reasoning in language models. arXiv preprint arXiv:2203.11171, 2022a.
  38. Self-instruct: Aligning language model with self generated instructions, 2022b.
  39. Finetuned language models are zero-shot learners. In International Conference on Learning Representations, 2022a. URL https://openreview.net/forum?id=gEZrGCozdqR.
  40. Chain-of-thought prompting elicits reasoning in large language models. Advances in Neural Information Processing Systems, 35:24824–24837, 2022b.
  41. React: Synergizing reasoning and acting in language models. arXiv preprint arXiv:2210.03629, 2022.
  42. Natural language is all a graph needs. arXiv preprint arXiv:2308.07134, 2023.
  43. Self-supervised contrastive cross-modality representation learning for spoken question answering. In Marie-Francine Moens, Xuanjing Huang, Lucia Specia, and Scott Wen-tau Yih, editors, Findings of the Association for Computational Linguistics: EMNLP 2021, pages 28–39, Punta Cana, Dominican Republic, November 2021. Association for Computational Linguistics. doi: 10.18653/v1/2021.findings-emnlp.3. URL https://aclanthology.org/2021.findings-emnlp.3.
  44. End-to-end spoken conversational question answering: Task, dataset and model. In Marine Carpuat, Marie-Catherine de Marneffe, and Ivan Vladimir Meza Ruiz, editors, Findings of the Association for Computational Linguistics: NAACL 2022, pages 1219–1232, Seattle, United States, July 2022. Association for Computational Linguistics. doi: 10.18653/v1/2022.findings-naacl.91. URL https://aclanthology.org/2022.findings-naacl.91.
  45. Mammoth: Building math generalist models through hybrid instruction tuning. arXiv preprint arXiv:2309.05653, 2023.
  46. Siren’s song in the ai ocean: A survey on hallucination in large language models. arXiv preprint arXiv:2309.01219, 2023.
  47. Graphtext: Graph reasoning in text space. arXiv preprint arXiv:2310.01089, 2023.
  48. Secrets of rlhf in large language models part i: Ppo. arXiv preprint arXiv:2307.04964, 2023.
  49. Least-to-most prompting enables complex reasoning in large language models. arXiv preprint arXiv:2205.10625, 2022.
User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (4)
  1. Nuo Chen (100 papers)
  2. Yuhan Li (49 papers)
  3. Jianheng Tang (31 papers)
  4. Jia Li (380 papers)
Citations (19)
View on Semantic Scholar