OmniInput: A Model-centric Evaluation Framework through Output Distribution

Weitang Liu, Ying Wai Li, Tianle Wang, Yi-Zhuang You, Jingbo Shang

Primary Area: general machine learning (i.e., none of the above)

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.

Keywords: comprehensive model evaluation; output distribution

Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2024/AuthorGuide.

TL;DR: Comprehensive model evaluation, entire input space.

Abstract: We propose a novel model-centric evaluation framework, OmniInput, to evaluate the quality of an AI/ML model's predictions on all possible inputs (including human-unrecognizable ones), which is crucial for AI safety and reliability. Unlike traditional data-centric evaluation based on pre-defined test sets, the test set in OmniInput is self-constructed by the model itself and the model quality is evaluated by investigating its output distribution. We employ an efficient sampler to obtain representative inputs and the output distribution of the trained model, which, after selective annotation, can be used to estimate the model's precision and recall at different output values and a comprehensive precision-recall curve. Our experiments demonstrate that OmniInput enables a more fine-grained comparison between models, especially when their performance is almost the same on pre-defined test sets, leading to new findings and insights for how to train more robust, generalizable models.

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: 748