Summary of Pairing Analogy-augmented Generation with Procedural Memory For Procedural Q&a, by K Roth and Rushil Gupta and Simon Halle and Bang Liu
Pairing Analogy-Augmented Generation with Procedural Memory for Procedural Q&A
by K Roth, Rushil Gupta, Simon Halle, Bang Liu
First submitted to arxiv on: 2 Sep 2024
Categories
- Main: Artificial Intelligence (cs.AI)
- Secondary: Computation and Language (cs.CL)
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 proposed formalism structures procedural knowledge to enable large language models (LLMs) to synthesize information into coherent plans for complex tasks. A novel dataset called LCStep is created from LangChain tutorials and leverages analogy-augmented generation (AAG), which uses a procedure memory store to adapt domain knowledge for new tasks. The approach outperforms few-shot and RAG baselines on three datasets, including LCStep, RecipeNLG, and CHAMP, under pairwise LLM-based evaluation and human evaluation. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Large language models can struggle with complex tasks that require planning. Researchers have developed a new way to organize information so machines can create plans for these tasks. They also created a dataset of tutorials to test this approach. To make it work, they used something called analogy-augmented generation (AAG), which helps machines adapt what they know from one task to solve another. The team tested AAG on several datasets and found that it did better than other approaches. |
Keywords
» Artificial intelligence » Few shot » Rag
