Summary of Nl2plan: Robust Llm-driven Planning From Minimal Text Descriptions, by Elliot Gestrin et al.
NL2Plan: Robust LLM-Driven Planning from Minimal Text Descriptions
by Elliot Gestrin, Marco Kuhlmann, Jendrik Seipp
First submitted to arxiv on: 7 May 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 The paper presents NL2Plan, a domain-agnostic offline Large Language Model (LLM)-driven planning system that combines the strengths of classical planners and LLMs. By incrementally extracting information from a short text prompt using an LLM, NL2Plan creates a complete PDDL description of both the domain and problem, which is then solved by a classical planner. The system is evaluated on four planning domains and shows a significant improvement over a plain chain-of-thought reasoning LLM approach, solving 10 out of 15 tasks. Additionally, NL2Plan reports failures instead of returning invalid plans, increasing explainability and making it an assistive tool for PDDL creation. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary NL2Plan is a new planning system that helps people create better plans by using language models to make the process easier and more accurate. Right now, planners can only work with specific types of input, which makes them hard to use. NL2Plan changes this by letting users give it any kind of text prompt, then using an LLM to extract the important information and turn it into a plan that a classical planner can solve. This system is tested on four different planning domains and shows big improvements over previous approaches, solving 10 out of 15 tasks. |
Keywords
» Artificial intelligence » Large language model » Prompt
