---
format: "aidr-story-markdown/v1"
id: "51888f116f5b083740bce79640ccacdc80055a08000c769931982fc83218805e"
canonical_url: "https://aidr.today/51888f11?lang=en"
title: "Turning Database Grading into RL Rewards: Post-Training Agents with ThinkingBox"
lang: "en"
requested_lang: "en"
available_langs: ["en","vi"]
translation_fallback: null
fallback_fields: []
published_at: "2026-10-22T22:24:34.000Z"
category: "Agents"
topics: ["microsoft","agent"]
source_urls: ["https://huggingface.co/blog/microsoft/thinkingbox-training"]
summary: "Microsoft has published a blog post on Hugging Face describing ThinkingBox, a method for turning database grading into reinforcement learning rewards for post-training agents."
---

# Turning Database Grading into RL Rewards: Post\-Training Agents with ThinkingBox

> [Open the canonical story](<https://aidr.today/51888f11?lang=en>)

**Published:** 2026-10-22T22:24:34.000Z
**Category:** Agents
**Topics:** microsoft, agent

## Summary

Microsoft has published a blog post on Hugging Face describing ThinkingBox, a method for turning database grading into reinforcement learning rewards for post\-training agents\.

## Sources

- [Story source](<https://huggingface.co/blog/microsoft/thinkingbox-training>)