Summary of Acpbench: Reasoning About Action, Change, and Planning, by Harsha Kokel et al.
ACPBench: Reasoning about Action, Change, and Planning
by Harsha Kokel, Michael Katz, Kavitha Srinivas, Shirin Sohrabi
First submitted to arxiv on: 8 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 The paper presents ACPBench, a benchmark for evaluating Large Language Models’ (LLMs) core planning skills. The benchmark consists of 7 reasoning tasks across 13 planning domains, allowing for automatic problem creation and scale. An evaluation of 22 LLMs and OpenAI o1 reveals a significant gap in the models’ reasoning capabilities, with notable gains in performance on multiple-choice questions but no progress on boolean questions. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper is about testing how well computer programs called Large Language Models can make decisions and solve problems that require planning. Right now, there are many of these models being used for things like making workflow decisions. To see if they’re good at this kind of thing, the researchers created a test with lots of different problems to solve. They found out that these models aren’t very good at solving some kinds of problems yet. |
