Skip to content

The ultimate harness for coding agents and beyond. All your agents. All your machines. One command center. Start with code, then follow your curiosity and build across disciplines: CAD, circuits, robots, games and music.

4.0 KB
Raw Blame
<section class="intro">
  <div>
    <p class="eyebrow">Experiment 024 / Synthetic materials lab</p>
    <h1>Find the signal.</h1>
    <p>
      Does a catalyst improve yield? Change the experiment, inspect each
      observation, and see what the data supports.
    </p>
  </div>
  <button id="export-csv" class="primary">Export observations ↗</button>
</section>
<main class="workspace">
  <div>
    <div class="metrics">
      <div class="metric">
        <strong id="sample-count"></strong><span>Visible samples</span>
      </div>
      <div class="metric">
        <strong id="effect-estimate"></strong><span>Observed uplift · pp</span>
      </div>
      <div class="metric">
        <strong id="r-squared"></strong><span>Visible fit · R²</span>
      </div>
    </div>
    <section class="panel chart-panel">
      <div class="panel-head">
        <span class="small">Temperature × reaction yield</span>
        <div class="actions">
          <button class="filter" id="control" aria-pressed="true">
            <i class="key control-key"></i>Control</button
          ><button class="filter" id="treatment" aria-pressed="true">
            <i class="key treatment-key"></i>Catalyst
          </button>
        </div>
      </div>
      <canvas
        id="chart"
        tabindex="0"
        role="img"
        aria-label="Scatter plot of temperature against yield. Exact observations are available in the table below."
      ></canvas>
      <p id="probe" class="probe mono" role="status">
        Hover a point to inspect an observation.
      </p>
    </section>
    <section class="inference">
      <div>
        <p class="eyebrow">What the experiment says</p>
        <h2 id="conclusion"></h2>
        <p class="small muted" id="interval"></p>
      </div>
      <canvas
        id="interval-chart"
        aria-label="Confidence interval for the observed treatment difference"
      ></canvas>
    </section>
    <details>
      <summary>
        Inspect observations <span class="muted">/ first 12 visible rows</span>
      </summary>
      <div class="table-scroll">
        <table>
          <thead>
            <tr>
              <th>Sample</th>
              <th>Group</th>
              <th>Temp · °C</th>
              <th>Yield · %</th>
            </tr>
          </thead>
          <tbody id="observations"></tbody>
        </table>
      </div>
    </details>
  </div>
  <aside class="side">
    <div>
      <p class="eyebrow">Design the experiment</p>
      <label class="field"
        ><span>True catalyst effect <output id="effect-value"></output></span
        ><input id="effect" type="range" min="-20" max="30" value="12" /></label
      ><label class="field"
        ><span>Measurement noise <output id="noise-value"></output></span
        ><input id="noise" type="range" min="1" max="20" value="9" /></label
      ><label class="field"
        ><span>Sample size <output id="count-value"></output></span
        ><input
          id="count"
          type="range"
          min="48"
          max="240"
          step="12"
          value="120" /></label
      ><button id="run" style="width: 100%">Replay collection</button>
    </div>
    <div class="section">
      <h3>A model you can inspect</h3>
      <p class="formula">
        yield = 28 + 0.48 × temperature<br />+ catalyst effect + noise
      </p>
      <p class="small muted" style="margin-top: 12px">
        Balanced groups, seeded temperatures, independent Gaussian measurement
        noise. The observed difference can depart from the true effect.
      </p>
    </div>
    <div class="section small muted">
      <strong>Synthetic data, by design.</strong>
      <p style="margin-top: 8px">
        The 95% interval uses a normal approximation for the unadjusted
        difference of means. It illustrates uncertainty; it is not a real
        experimental finding.
      </p>
    </div>
  </aside>
</main>
<footer class="status-line">
  <span><i class="dot"></i>REPRODUCIBLE OBSERVATIONS</span
  ><span>SEED / PROBE / QUESTION / REPEAT</span>
</footer>