Predicting from Strings: Language Model Embeddings for Bayesian Optimization
Keywords: language, model, embedding, bayesian, optimization, in, context, regression, gaussian, process, meta, learning
TL;DR: "Embed-then-Regress": Use language models to embed strings as features, to be sent into Transformer neural processes for Bayesian Optimization.
Abstract: Bayesian Optimization is ubiquitous in the field of experimental design and blackbox optimization for improving search efficiency, but has been traditionally restricted to regression models which are only applicable to fixed search spaces and tabular input features. We propose _Embed-then-Regress_, a paradigm for applying in-context regression over string inputs, through the use of string embedding capabilities of pretrained language models. By expressing all inputs as strings, we able to perform general-purpose regression for Bayesian Optimization over different search domains such as traditional and combinatorial optimization, obtaining comparable results to state-of-the-art Gaussian Process-based algorithms.
Primary Area: foundation or frontier models, including LLMs
Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.
Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2025/AuthorGuide.
Reciprocal Reviewing: I understand the reciprocal reviewing requirement as described on https://iclr.cc/Conferences/2025/CallForPapers. If none of the authors are registered as a reviewer, it may result in a desk rejection at the discretion of the program chairs. To request an exception, please complete this form at https://forms.gle/Huojr6VjkFxiQsUp6.
Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.
No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.
Submission Number: 5026
Loading