Summary of Graph-enhanced Large Language Models in Asynchronous Plan Reasoning, by Fangru Lin et al.
Graph-enhanced Large Language Models in Asynchronous Plan Reasoning
by Fangru Lin, Emanuele La Malfa, Valentin Hofmann, Elle Michelle Yang, Anthony Cohn, Janet B. Pierrehumbert
First submitted to arxiv on: 5 Feb 2024
Categories
- Main: Artificial Intelligence (cs.AI)
- Secondary: Computation and Language (cs.CL); Machine Learning (cs.LG)
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 study, researchers investigate whether large language models (LLMs) can succeed in planning tasks that require sequential and parallel processing. They present a benchmark called AsyncHow and find that representative LLMs, including GPT-4 and LLaMA-2, perform poorly when not provided with illustrations of the task-solving process. To overcome this limitation, they propose Plan Like a Graph (PLaG), a novel technique combining graphs with natural language prompts, which achieves state-of-the-art results. However, they also observe that while PLaG can improve model performance, LLMs still degrade significantly when task complexity increases, highlighting their limitations in simulating digital devices. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Large language models are trying to learn how to plan and think ahead like humans do. The problem is that this kind of planning requires doing things one after another, but also at the same time. Can machines really do this? In a new study, scientists test some top language models and find that they don’t do very well when they’re not given hints about how to solve the problem. To help these models, they create a new way of thinking called Plan Like a Graph. This helps them do better, but there’s still a limit to what they can do. |
Keywords
* Artificial intelligence * Gpt * Llama
