# Brittleness vs sample efficiency

**URL:** <https://discourse.numenta.org/t/brittleness-vs-sample-efficiency/10250>\
**Category:** Lounge\
**Created:** [November 8, 2022, 11:19am UTC](https://discourse.numenta.org/t/brittleness-vs-sample-efficiency/10250 "2022-11-08T11:19:33Z")\
**Posts on this page:** 1\
**Showing post:** 6

<div class="post-metadata">

**Author:** ![cezar\_t](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.numenta.org/cezar_t/32/4451_2.png) [@cezar\_t](https://discourse.numenta.org/u/cezar_t)\
**Post date:** [November 11, 2022, 7:59pm UTC](https://discourse.numenta.org/t/brittleness-vs-sample-efficiency/10250/6 "2022-11-11T19:59:57Z")

</div>

Sample efficiency can also be viewed as number of datapoints needed to solve the problem.  
In cartpole case, this means the number of environment observations and with [this encoder](https://discourse.numenta.org/t/scalar-vectors-as-intermediate-stages/10259) sometimes is lower than 30

Here-s the updated code - with both “robust” and “brittle” solutions. [GitHub - Blimpyway/CartPoleChallenge: Balancing the cartpole. Fast](https://github.com/Blimpyway/CartPoleChallenge)

---

_[View the full topic](https://discourse.numenta.org/t/brittleness-vs-sample-efficiency/10250)._