Summary of Graphic: a Graph-based In-context Example Retrieval Model For Multi-step Reasoning, by Jiale Fu et al.
GraphIC: A Graph-Based In-Context Example Retrieval Model for Multi-Step Reasoning
by Jiale Fu, Yaqing Wang, Simeng Han, Jiaming Fan, Xu Yang
First submitted to arxiv on: 3 Oct 2024
Categories
- Main: Artificial Intelligence (cs.AI)
- Secondary: None
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This paper proposes GraphIC, a graph-based retrieval model that leverages reasoning-aware representation and specialized similarity metric for in-context example retrieval. Current methods use text embeddings to measure semantic similarity, which can introduce bias in multi-step reasoning tasks. GraphIC constructs thought graphs-directed, node-attributed graphs that explicitly model reasoning steps and their dependencies-for candidate examples and queries. This approach filters out superficial semantics while preserving essential reasoning processes. Next, GraphIC retrieves examples using a novel similarity metric tailored for these graphs, capturing sequential reasoning patterns and asymmetry between examples. The results show that GraphIC outperforms 10 baseline methods in comprehensive evaluations across mathematical reasoning, code generation, and logical reasoning tasks. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper helps make language models better by letting them learn from examples related to a specific task or problem. Right now, this process can be limited because the model doesn’t understand how the example relates to the task. The authors propose a new way of finding examples that takes into account how they are connected and what reasoning is required to solve the problem. This method is tested on different tasks like math problems, code writing, and logic puzzles, and it performs better than other methods. |
Keywords
» Artificial intelligence » Semantics
