QualEval: Qualitative Evaluation for Model Improvement

ICLR 2024 Workshop DMLR Submission14 Authors

Keywords: Qualitative framework for model-improvement, structured optimization framework, data-scientist-in-a-box

TL;DR: We propose QualEval, the first qualitative evaluation framework for LLMs that automatically provides natural language actionable insights to improve the model performance.

Abstract:

Quantitative evaluation metrics have been pivotal in gauging the advancements of AI systems like large language models (LLMs).
However, due to the intricate nature of real-world tasks, a single scalar to quantify and compare performance trivializes the fine-grained nuances of model behavior. Additionally, metrics do not yield actionable diagnostics for model improvement, thus requiring extensive manual efforts of scientists, involving sifting through vast datasets and attempting hit-or-miss adjustments to training data or setups. In this work, we address the shortcomings of quantitative metrics by proposing QualEval, which uses automated qualitative evaluation as a vehicle for model improvement. QualEval uses a powerful LLM reasoner and our novel flexible linear programming solver to generate human-readable insights that when applied, accelerate model improvement. The insights are supported by a dashboard report with fine-grained visualizations and human-interpretable analyses. We corroborate the faithfulness of QualEval by demonstrating that leveraging its insights, for example, improves the absolute performance of the Llama 2 model by up to 15\% points relative on a challenging dialogue task (DialogSum) when compared to baselines. QualEval successfully increases the pace and quality of model development by eliminating the need of arduous manual analysis, thus serving as a data-scientist-in-a-box.

Primary Subject Area: Data-centric explainable AI

Paper Type: Research paper: up to 8 pages

Participation Mode: In-person

Confirmation: I have read and agree with the workshop's policy on behalf of myself and my co-authors.

Submission Number: 14