AgentArena / templates /matrix.html
smlflg's picture
Initial public upload from Projekte/AgentArena
3b0d6b3 verified
Raw History Blame Contribute Delete
3.69 kB
{% extends "base.html" %}
{% block title %}Harness x Model Matrix — AgentArena{% endblock %}
{% block content %}
<div class="hero" style="padding-bottom:16px;">
<h1>Harness x Model Matrix</h1>
<p>Same LLM, different harnesses — who works cleaner, faster, cheaper?</p>
</div>
{% if harnesses and models %}
<div class="table-wrapper mb-24">
<table>
<thead>
<tr>
<th>Harness</th>
{% for model in models %}
<th style="text-align:center;">{{ model }}</th>
{% endfor %}
</tr>
</thead>
<tbody>
{% for harness in harnesses %}
<tr>
<td>
<a href="/harness/{{ harness }}"><strong>{{ harness }}</strong></a>
</td>
{% for model in models %}
{% set key = harness ~ "|" ~ model %}
{% set cell = cells.get(key) %}
<td style="text-align:center; padding:8px 12px;">
{% if cell %}
<div style="margin-bottom:4px;">
<div class="score-bar" style="justify-content:center;">
<span style="min-width:40px; font-weight:700; {% if cell.pass_at_1 >= 0.7 %}color:var(--success){% elif cell.pass_at_1 >= 0.4 %}color:var(--warning){% else %}color:var(--error){% endif %}">
{{ "%.0f"|format(cell.pass_at_1 * 100) }}%
</span>
</div>
<div style="font-size:10px; color:var(--muted);">pass@1</div>
</div>
{% if cell.avg_composite is not none %}
<div style="margin-bottom:2px;">
<span class="mono" style="font-size:12px; {% if cell.avg_composite >= 0.7 %}color:var(--success){% elif cell.avg_composite >= 0.4 %}color:var(--warning){% else %}color:var(--error){% endif %}">
{{ "%.0f"|format(cell.avg_composite * 100) }}%
</span>
<span style="font-size:10px; color:var(--muted);">comp</span>
</div>
{% endif %}
<div style="font-size:11px; color:var(--muted);">
${{ "%.4f"|format(cell.avg_cost) }}
<span style="margin-left:4px;">{{ "%.0f"|format(cell.avg_duration) }}s</span>
<span style="margin-left:4px;">{{ cell.total_runs }}r</span>
</div>
{% else %}
<span class="text-muted" style="font-size:18px;">—</span>
{% endif %}
</td>
{% endfor %}
</tr>
{% endfor %}
</tbody>
</table>
</div>
<!-- Legend -->
<div class="card" style="padding:16px 20px;">
<div style="font-size:12px; color:var(--muted); text-transform:uppercase; letter-spacing:0.5px; margin-bottom:8px;">
Legend
</div>
<div style="display:flex; gap:24px; flex-wrap:wrap; font-size:13px; color:var(--muted);">
<span><strong style="color:var(--success);">70%+</strong> = strong</span>
<span><strong style="color:var(--warning);">40-69%</strong> = moderate</span>
<span><strong style="color:var(--error);">&lt;40%</strong> = weak</span>
<span><strong>pass@1</strong> = reliability</span>
<span><strong>comp</strong> = composite score</span>
<span><strong>r</strong> = total runs</span>
</div>
</div>
{% else %}
<div class="card">
<div class="empty">
<h3>No data yet</h3>
<p>Run benchmarks with different harness/model combinations to populate the matrix.</p>
<a href="/" class="btn btn-primary">Go to Leaderboard</a>
</div>
</div>
{% endif %}
<!-- Harness Steckbriefe -->
{% if harnesses %}
<h2 style="margin-top:32px;">Harness Reports</h2>
<div style="display:flex; gap:12px; flex-wrap:wrap;">
{% for harness in harnesses %}
<a href="/harness/{{ harness }}" class="btn btn-secondary">{{ harness }}</a>
{% endfor %}
</div>
{% endif %}
{% endblock %}