Summary of Lost-in-distance: Impact Of Contextual Proximity on Llm Performance in Graph Tasks, by Hamed Firooz et al.
Lost-in-Distance: Impact of Contextual Proximity on LLM Performance in Graph Tasks
by Hamed Firooz, Maziar Sanjabi, Wenlong Jiang, Xiaoling Zhai
First submitted to arxiv on: 2 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 In this paper, researchers investigate the limitations of Large Language Models (LLMs) in processing complex graph data. They find that LLMs struggle when retrieving contextual information from distant parts of a graph, which they term “lost-in-distance”. The study examines two graph tasks and shows how model performance is affected by the proximity of relevant edges within the graph. The authors evaluate three publicly available LLMs using various encoding techniques and propose a formulation for the lost-in-distance phenomenon. Their results indicate that model accuracy can decline significantly as distance between node connections increases, regardless of graph encoding or model size. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This study looks at how well Large Language Models (LLMs) work with complex graphs. Graphs are like big pictures made up of nodes and edges. The researchers found that LLMs have trouble finding important information in these graphs if it’s far away from the start. They tested three different LLMs to see how well they did on two tasks: identifying connections between nodes, and seeing how similar certain nodes are. The results show that as the distance between nodes increases, the model’s accuracy can drop by a lot. |
