Summary of Bootstrapping Linear Models For Fast Online Adaptation in Human-agent Collaboration, by Benjamin a Newman et al.
Bootstrapping Linear Models for Fast Online Adaptation in Human-Agent Collaboration
by Benjamin A Newman, Chris Paxton, Kris Kitani, Henny Admoni
First submitted to arxiv on: 16 Apr 2024
Categories
- Main: Artificial Intelligence (cs.AI)
- Secondary: Human-Computer Interaction (cs.HC); Robotics (cs.RO)
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 This paper proposes a novel approach to initializing policies for agents that assist people, allowing them to adapt quickly to align with their partners’ reward functions. The method, called BLR-HAC (Bootstrapped Logistic Regression for Human Agent Collaboration), combines the benefits of large nonlinear models and low-capacity models to achieve efficient online learning. By bootstrapping large nonlinear models to learn the parameters of a low-capacity model, BLR-HAC enables rapid inference and fine-tuning updates during collaboration. The approach is tested in a simulated surface rearrangement task, demonstrating higher zero-shot accuracy than shallow methods while requiring less computation to adapt online. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper helps machines work better with people by learning how to quickly adjust to the person’s goals. The researchers created a new method called BLR-HAC that combines two different types of models to make this happen. It starts with a big model and then adjusts it using a smaller model, which allows it to learn quickly and accurately while still being efficient with its computations. |
Keywords
» Artificial intelligence » Bootstrapping » Fine tuning » Inference » Logistic regression » Online learning » Zero shot
