# What’s the best way to evaluate model hallucination rates in generative models?

**URL:** <https://community.deeplearning.ai/t/what-s-the-best-way-to-evaluate-model-hallucination-rates-in-generative-models/857382>\
**Category:** AI Discussions\
**Tags:** ai-discussions\
**Created:** [July 22, 2025, 9:32am UTC](https://community.deeplearning.ai/t/what-s-the-best-way-to-evaluate-model-hallucination-rates-in-generative-models/857382 "2025-07-22T09:32:28Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![Marcus10](https://sea1.discourse-cdn.com/dlai/user_avatar/community.deeplearning.ai/marcus10/32/115251_2.png) [@Marcus10](https://community.deeplearning.ai/u/Marcus10)\
**Post date:** [July 22, 2025, 9:32am UTC](https://community.deeplearning.ai/t/what-s-the-best-way-to-evaluate-model-hallucination-rates-in-generative-models/857382/1 "2025-07-22T09:32:28Z")

</div>

Hi everyone,

I’m currently working with a generative language model and exploring ways to measure how often it hallucinates (i.e., generates factually incorrect or made-up content). While human evaluation is one approach, I’m looking for more scalable, automated metrics or frameworks that are reliable.

What methods or tools have you found effective for hallucination detection? Are there any benchmarks or best practices you’d recommend for evaluating factual consistency in outputs?

---

<div class="post-metadata">

**Author:** ![TMosh](https://sea1.discourse-cdn.com/dlai/user_avatar/community.deeplearning.ai/tmosh/32/9321_2.png) [@TMosh](https://community.deeplearning.ai/u/TMosh)\
**Post date:** [July 22, 2025, 5:27pm UTC](https://community.deeplearning.ai/t/what-s-the-best-way-to-evaluate-model-hallucination-rates-in-generative-models/857382/2 "2025-07-22T17:27:49Z")

</div>

I’m not aware of any automated tool that can reliably check the work of another automated tool.

---

<div class="post-metadata">

**Author:** ![Marcus10](https://sea1.discourse-cdn.com/dlai/user_avatar/community.deeplearning.ai/marcus10/32/115251_2.png) [@Marcus10](https://community.deeplearning.ai/u/Marcus10)\
**Post date:** [August 7, 2025, 11:12am UTC](https://community.deeplearning.ai/t/what-s-the-best-way-to-evaluate-model-hallucination-rates-in-generative-models/857382/3 "2025-08-07T11:12:25Z")

</div>

Ok