> ## Documentation Index
> Fetch the complete documentation index at: https://docs.nora.my/llms.txt
> Use this file to discover all available pages before exploring further.

# New experiment

> Select a cluster, configure options, launch. Three steps and it runs

Experiments start in two places: **New experiment** at the top right of Simulation, or **Run simulation** on a cluster's detail page. Either way the same drawer opens, and after three steps the experiment runs in the background.

<Frame>
  <img src="https://mintcdn.com/conscience-technology/oQkz_vdmMGZlOMYc/images/docs/simulation-new-experiment.png?fit=max&auto=format&n=oQkz_vdmMGZlOMYc&q=85&s=42c06795be15790d89f08c84f2357fbb" alt="Step 2 of the new-experiment drawer, showing candidate count, budget, the three experiment options, and the suggestions box" width="1512" height="809" data-path="images/docs/simulation-new-experiment.png" />
</Frame>

<Steps>
  <Step title="Select cluster">
    "Which cluster should we fix?" Pick a cluster from the list. Each row carries its category tag and signal count, and **Search clusters** narrows it down. If you came from a cluster's detail page, this step is already filled in.
  </Step>

  <Step title="Configure options">
    "Which options should we use?"

    * **Candidates** 2, 3, 4, or 6. The default is 4.
    * **Budget (USD)** the spending cap for the experiment. The default is \$2.
    * **Experiment options**
      * **Cost optimization** sweeps models and config for a cheaper candidate that keeps accuracy.
      * **Consistency** replays each input k times and measures stability. Turning it on lets you set k (default 3).
      * **Hold-out regression** guards previously-correct cases against regression. Choose Off, Light, or Full.
    * **Suggestions for this experiment (optional)** free text to steer the direction. For example: "Keep the prompt as-is and try improving it by tweaking guardrails, memory, or retrieval instead."
    * **More options → Reference folders (optional)** pick folders to score candidates against a hand-built golden set. If you pick none, the cluster's own cases are used.
  </Step>

  <Step title="Launch">
    "Shall we start?" The cluster, category, candidates, budget, holdout, and instructions are summarized. Press **Start experiment**.
  </Step>
</Steps>

## While it runs

The experiment runs in the background. A measuring card appears in the board's **Experiment** column, and opening it shows candidates being built in real time (see [Experiment results](/reliable/simulation/results)). You can close the drawer and do something else.

## Learn more

<CardGroup cols={2}>
  <Card title="Experiment results" icon="chart-line" href="/reliable/simulation/results">
    Read the candidates and metrics the experiment produced.
  </Card>

  <Card title="Guarding against regressions" icon="shield" href="/reliable/simulation/regressions">
    What the Hold-out regression option does.
  </Card>
</CardGroup>