Summary of Maple: a Framework For Active Preference Learning Guided by Large Language Models, By Saaduddin Mahmud et al.
MAPLE: A Framework for Active Preference Learning Guided by Large Language Models
by Saaduddin Mahmud, Mason Nakamura, Shlomo Zilberstein
First submitted to arxiv on: 10 Dec 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); 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 paper introduces MAPLE, a framework for large language model-guided Bayesian active preference learning, which leverages LLMs to model the distribution over preference functions while conditioning on natural language feedback. The approach employs active learning to reduce uncertainty and incorporates a query selection mechanism to identify informative and easy-to-answer queries. Evaluations across two benchmarks, including a real-world vehicle route planning benchmark using OpenStreetMap data, demonstrate MAPLE’s sample efficiency and preference inference quality. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary MAPLE is a new way to learn people’s preferences by using big language models. Currently, this process can be slow, require lots of human supervision, and be hard to understand. MAPLE makes it faster, more efficient, and easier to interpret by combining natural language feedback with traditional preference learning methods. It also helps reduce the amount of work humans need to do by identifying the most important questions to ask. The results show that MAPLE can quickly learn people’s preferences and make good decisions. |
Keywords
» Artificial intelligence » Active learning » Inference » Large language model
