Summary of Reinforced In-context Black-box Optimization, by Lei Song et al.
Reinforced In-Context Black-Box Optimization
by Lei Song, Chenxiao Gao, Ke Xue, Chenyang Wu, Dong Li, Jianye Hao, Zongzhang Zhang, Chao Qian
First submitted to arxiv on: 27 Feb 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); Neural and Evolutionary Computing (cs.NE)
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 method, RIBBO, is a meta-learning approach for optimizing Black-Box Optimization (BBO) algorithms. By learning an end-to-end model from offline data, RIBBO can automatically generate query points that satisfy user-desired regret without requiring tedious hand-crafted heuristics. The method leverages large sequence models to learn optimization histories and extract task information, and incorporates “regret-to-go” tokens to represent cumulative regret over future parts of the histories. Empirical results demonstrate RIBBO’s universality on diverse problems, including BBO benchmark functions, hyper-parameter optimization, and robot control problems. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary RIBBO is a new way to optimize algorithms without needing to make decisions about what works best. It learns from past experiences and can choose good options automatically. This helps solve complex problems like finding the right settings for robots or computers. RIBBO uses special tokens that help it understand how well an algorithm will do in the future, based on its past performance. This approach has shown to work well across many different types of problems. |
Keywords
* Artificial intelligence * Meta learning * Optimization
