Graphs and LLMs
There are a lot of graph problems being re-thought __for LLMs__ and __by LLMs__
February 11,2024
last update: 2024-11-19
I’ll add the ones with the most fun:utilization first
Building Graphs using LLMs:
This strategy comes clearly under the by LLMs category and the most obvious one as well. Because of their semantic understanding, LLMs can be prompted into providing the triplets of relationship between the entities in a piece of text. Throw a piece of text, and it draws a graph of the relationship between subjects and objects This is a relatively straightforward utilization of an LLM’s ability to find pattern in text. eg. GraphGPTw
Pathfinding in a Graph
LLMs are not trained to traverse graphs. Especially, they’ve not been explicitly taught how to model a graph from a triplet. But seems like they perform fairly well. The first graph in the blog from [jacob brazeal]https://jacobbrazeal.wordpress.com/2022/09/23/gpt-3-can-find-paths-up-to-7-nodes-long-in-random-graphs/) is very telling. GPT-3 was able to