Summary of Agent Planning with World Knowledge Model, by Shuofei Qiao et al.
Agent Planning with World Knowledge Model
by Shuofei Qiao, Runnan Fang, Ningyu Zhang, Yuqi Zhu, Xiang Chen, Shumin Deng, Yong Jiang, Pengjun Xie, Fei Huang, Huajun Chen
First submitted to arxiv on: 23 May 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG); Multiagent Systems (cs.MA)
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 abstract discusses recent advancements in using large language models (LLMs) as agent models for interactive planning tasks. While LLMs have shown promising results, they still struggle with trial-and-error and hallucinatory actions due to their lack of understanding of the physical world. The authors introduce a parametric World Knowledge Model (WKM) that provides prior task knowledge for global planning and dynamic state knowledge for local planning. Experiments on three complex datasets using three LLMs demonstrate superior performance compared to baselines, alleviating blind trial-and-error and hallucinatory actions. Interesting findings include generalization of instance-level task knowledge, weak WKM guiding strong agent models, and unified WKM training with potential for further development. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper uses large language models as agents to plan interactive tasks. It shows that these models can do well, but they struggle because they don’t understand the real world. To fix this, the authors create a “world knowledge model” that helps the agent make better plans. They test their idea on three big datasets and it works really well! The results show that the agent makes fewer mistakes and does a better job of planning. |
Keywords
» Artificial intelligence » Generalization
