• Perle Labs Secures $17.5M in Funding
  • Season 1 is live!
  • Perle Labs Secures $17.5M in Funding
  • Season 1 is live!
  • Perle Labs Secures $17.5M in Funding
  • Join The Beta Launch
  • Perle Labs Secures $17.5M in Funding
  • Join The Beta Launch
  • Perle Labs Secures $17.5M in Funding
  • Join The Beta Launch
  • Perle Labs Secures $17.5M in Funding
  • Join The Beta Launch
This new task invites the community to apply human judgment to separate facts from opinions, improving how AI systems learn from language.

We’re rolling out a new task on Perle Labs, and this one’s all about attention to detail.

FactHunt Quest invites the community to help AI learn how to separate verifiable facts from opinions, emotions, and assumptions. Every paragraph in this task includes at least one factual claim, and some include two. It is your job to identify them.

Why We Built This Task

In real-world text, facts are rarely presented on their own. They’re usually wrapped in opinions, impressions, or personal reactions.

FactHunt is designed to reflect that reality. By relying on human contributors to identify factual claims precisely, this task helps build datasets that are clearer, more reliable, and far more useful for training AI systems that need to reason accurately.

What You’ll Do

You’ll be given short paragraphs made up of several sentences. Each paragraph includes one or more factual statements mixed in with opinions or subjective language.

Your goal is to identify every sentence that contains a verifiable factual claim.

For each paragraph:

  • Read the full passage first to understand the context
  • Go sentence by sentence and identify a factual claim
  • Highlight the sentence and click Submit (include the punctuation)
  • Repeat this process until all factual sentences in the paragraph have been identified

Every paragraph includes at least one correct answer, and some include more than one. The task will guide you on how many factual sentences there are, so take your time to identify them all.

You’ll see feedback immediately after each submission.

Why Your Contribution Matters

This task relies on human judgment to help AI recognize the patterns and phrasing that signal factual statements, as opposed to opinions, emotions, or speculation. Those cues matter, because they shape how models interpret text in real-world contexts.

By contributing, you’re helping build datasets that reflect how people actually communicate facts, leading to AI systems that are clearer, more reliable, and easier to trust.

Ready to Get Started?

FactHunt Quest is now live on your Perle Labs dashboard here.

If you’re already in beta, you can start right away by heading to the app and picking up the task. Each annotation you submit helps improve how AI systems learn to recognize and reason over factual information in real-world language.

As always, keep an eye on X, Discord, and Telegram for updates on new tasks, platform improvements, and ways to get involved with the Perle Labs community.

Wishing you all a happy holiday season and a wonderful start to the new year!

New Task is Live on Perle Labs Beta: FactHunt Quest