Get the latest tech news

Achieving 10,000x training data reduction with high-fidelity labels


7, 2025 Markus Krause, Engineering Manager, and Nancy Chang, Research Scientist, Google Ads A new active learning method for curating high-quality data that reduces training data requirements for fine-tuning LLMs by orders of magnitude. Classifying unsafe ad content has proven an enticing problem space for leveraging large language models (LLMs).

The inherent complexity involved in identifying policy-violating content demands solutions capable of deep contextual and cultural understanding, areas of relative strength for LLMs over traditional machine learning systems. With this in mind, we describe a new, scalable curation process for active learning that can drastically reduce the amount of training data needed for fine-tuning LLMs while significantly improving model alignment with human experts. But given sufficient label quality, our curation process is able to leverage the strengths of both LLMs, which can cast a wide net over the problem space, and domain experts, who can focus more efficiently on the most challenging examples.

Get the Android app

Or read this on Hacker News

Read more on:

Photo of fidelity labels

fidelity labels