Summary from the original article
Asking a language model to pick a category usually ends with code that parses its prose, hopes the label is spelled the way you expected and breaks the day the model adds a polite sentence before the answer. When I built Linkmortem , I solved that by asking Gemma 4 for a single JSON object with a fixed list of page kinds and validating it in code. Microsoft Foundry now offers a more direct route for this kind of question,…
By Carlos José Castro Galante ·
AI key points, how left, center and right framed this story, tone by side and loaded language are part of News.gg Premium.
Upgrade this serverAsking a language model to pick a category usually ends with code that parses its prose, hopes the label is spelled the way you expected and breaks the day the model adds a polite sentence before the answer. When I built Linkmortem , I solved that by asking Gemma 4 for a single JSON object with a fixed list of page kinds and validating it in code. Microsoft Foundry now offers a more direct route for this kind of question, the Azure OpenAI Decisions API, which returns typed answers instead of generated text. This post is for Python developers who classify, filter or route text on Azure and…
Yesterday · dev.to · You are reading this one
dev.to · 3 hours ago
Similarity Is Not Truth: Comparing vector distance, is_true(), and extract() in Azure AI for PostgreSQLdev.to · 23 hours ago
How to write expected behavior for LLM test cases