Some Practical Concerns and Solutions for Using Pretrained Representation in Industrial Systems

Da Xu

Keywords: Representation Learning, Stability, Generalization, Convergence, Predictability, Industry Application

Abstract:

Deep learning has dramatically changed the way data scientists and engineers craft features -- the once tedious process of measuring and constructing can now be achieved by training learnable representations. Recent work shows pretraining can endow representations with relevant signals, and in practice they are often used as feature vectors in downstream models. In real-world production, however, we have encountered key problems that cannot be justified by existing knowledge. They raise concerns that the naive use of pretrained representation as feature vector could lead to unwarranted and suboptimal solution.
Our investigation reveals critical insights into the gap of uniform convergence for analyzing pretrained representations, their stochastic nature under gradient descent optimization, what does model convergence means to them, and how they might interact with downstream tasks. Inspired by our analysis, we explore a simple yet powerful approach that can refine pretrained representation in multiple ways, which we call "Featurizing Pretrained Representations". Our work balances practicality and rigor, and contributes to both applied and theoretical research of representation learning.

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.

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: Yes

Please Choose The Closest Area That Your Submission Falls Into: Deep Learning and representational learning

TL;DR: We investigate some practical concerns and solutions for using pretrained representation in industrial systems.