Summary of Prise: Llm-style Sequence Compression For Learning Temporal Action Abstractions in Control, by Ruijie Zheng et al.
PRISE: LLM-Style Sequence Compression for Learning Temporal Action Abstractions in Control
by Ruijie Zheng, Ching-An Cheng, Hal Daumé III, Furong Huang, Andrey Kolobov
First submitted to arxiv on: 16 Feb 2024
Categories
- Main: Machine Learning (cs.LG)
- 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 proposed novel view treats inducing temporal action abstractions as a sequence compression problem, bringing input tokenization via byte pair encoding (BPE) to the task of learning skills of variable time span in continuous control domains. The Primitive Sequence Encoding (PRISE) approach combines continuous action quantization with BPE to learn powerful action abstractions. Experimental results show that high-level skills discovered by PRISE from a multitask set of robotic manipulation demonstrations significantly boost performance in both multitask imitation learning and few-shot imitation learning on unseen tasks. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper proposes a new way to understand and create rules for robots to follow in different situations. They use an old technique called byte pair encoding (BPE) that is used in language models, but apply it to robot actions instead of words. This helps robots learn how to do complex things by breaking them down into simpler steps. The approach was tested on a set of robotic manipulation tasks and showed great results. |
Keywords
* Artificial intelligence * Few shot * Quantization * Tokenization
