arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2406.09657cs.LGstat.ML
收藏

使用LES缓解潜在空间优化中的过度探索问题

Mitigating over-exploration in latent space optimization using LES

UC Berkeley(加州大学伯克利分校);Rice University(莱斯大学);Brown University(布朗大学)

Omer Ronen, Ahmed Imtiaz Humayun, Richard Baraniuk, Randall Balestriero, Bin Yu

AI总结:

针对潜在空间优化(LSO)的过度探索问题,提出无需额外训练的潜在探索分数(LES),经多基准多模型验证可提升解质量,为LSO研究开辟新路径。

摘要:

我们提出了潜在探索分数(Latent Exploration Score,LES),用于缓解潜在空间优化(Latent Space Optimization,LSO)中的过度探索问题——LSO是求解黑盒离散优化问题的常用方法。LSO在变分自编码器(Variational Autoencoder,VAE)的潜在空间内执行连续优化,已知其易受过度探索问题影响,该问题会产生不切实际的解,降低方法的实用性。LES利用已训练解码器对数据分布的近似,可适配任意VAE解码器(包括预训练解码器),无需额外训练、修改架构或访问训练数据。我们在5项LSO基准任务和22个VAE模型上开展评估,结果表明LES始终能提升解的质量,同时保持较高的目标值,在多数情况下优于现有方案。LES具备分布外区域识别能力、可微性与计算易处理性,我们认为这将为LSO研究开辟新方向。LES的开源代码可在https://github.com/OmerRonen/les获取。

英文摘要:

We develop Latent Exploration Score (LES) to mitigate over-exploration in Latent Space Optimization (LSO), a popular method for solving black-box discrete optimization problems. LSO utilizes continuous optimization within the latent space of a Variational Autoencoder (VAE) and is known to be susceptible to over-exploration, which manifests in unrealistic solutions that reduce its practicality. LES leverages the trained decoder's approximation of the data distribution, and can be employed with any VAE decoder - including pretrained ones - without additional training, architectural changes or access to the training data. Our evaluation across five LSO benchmark tasks and twenty-two VAE models demonstrates that LES always enhances the quality of the solutions while maintaining high objective values, leading to improvements over existing solutions in most cases. We believe that new avenues to LSO will be opened by LES' ability to identify out of distribution areas, differentiability, and computational tractability. Open source code for LES is available at https://github.com/OmerRonen/les.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑