repro-bundle-v2/repro/poster_embed.html

<!-- poster_embed.html -->
<div style="font-family:system-ui,-apple-system,Segoe UI,sans-serif;background:#0f172a;color:#e2e8f0;padding:22px;border-radius:12px;line-height:1.45">
  <h2 style="color:#38bdf8;margin:0 0 4px">Reproduction: DDSVM - A Differentiable Framework for Deep Support Vector Machines with Iterative Geometry-Aware Optimization</h2>
  <p style="margin:0 0 14px;color:#94a3b8;font-size:.92em">ICML 2026 submission #34342 (OpenReview <code>l6MbbwsWUs</code>) &middot; CPU-only reproduction &middot; 4 datasets x 6 methods x 25 seeds &middot; 15.7 min total compute</p>

  <div style="display:grid;grid-template-columns:repeat(3,1fr);gap:12px;margin-bottom:14px">
    <div style="background:#1e293b;padding:12px;border-radius:8px;border-left:4px solid #fbbf24">
      <div style="color:#fbbf24;font-weight:600;margin-bottom:4px">Claim 1 - alternating optimization</div>
      <div style="font-size:1.6em;font-weight:700">PARTIAL</div>
      <div style="font-size:.85em;color:#cbd5e1">7/12 pre-stated checks. Three-phase block structure is exact (cross-phase parameter drift = 0.0 in all 1500 cycle-observations), but log-linear convergence holds on only 1/4 datasets.</div>
    </div>
    <div style="background:#1e293b;padding:12px;border-radius:8px;border-left:4px solid #fbbf24">
      <div style="color:#fbbf24;font-weight:600;margin-bottom:4px">Claim 2 - push along the normal</div>
      <div style="font-size:1.6em;font-weight:700">PARTIAL</div>
      <div style="font-size:.85em;color:#cbd5e1">9/16. Margin grows and the support set shrinks on 4/4 datasets, but the geometry-aware push buys <b>no</b> test margin over a random-direction push (4/4 CIs straddle 0).</div>
    </div>
    <div style="background:#1e293b;padding:12px;border-radius:8px;border-left:4px solid #fbbf24">
      <div style="color:#fbbf24;font-weight:600;margin-bottom:4px">Claim 3 - beats baselines</div>
      <div style="font-size:1.6em;font-weight:700">PARTIAL</div>
      <div style="font-size:.85em;color:#cbd5e1">Beats both deep baselines on 1/4 datasets (gauss-xor: +4.39pp vs CE, +3.00pp vs deep-SVM). An off-the-shelf RBF SVM still wins there (-0.84pp).</div>
    </div>
  </div>

  <table style="width:100%;border-collapse:collapse;font-size:.86em;background:#1e293b;border-radius:8px;overflow:hidden">
    <thead><tr style="background:#334155;color:#f1f5f9">
      <th style="padding:6px 8px;text-align:left">dataset</th>
      <th style="padding:6px 8px;text-align:right">DDSVM acc % (95% CI)</th>
      <th style="padding:6px 8px;text-align:right">vs deep-CE (pp)</th>
      <th style="padding:6px 8px;text-align:right">vs deep-SVM (pp)</th>
      <th style="padding:6px 8px;text-align:right">vs RBF-SVM (pp)</th>
    </tr></thead>
    <tbody><tr><td style='padding:4px 8px;border-bottom:1px solid #334155'>moons-hard</td><td style='padding:4px 8px;border-bottom:1px solid #334155;text-align:right'>91.16 +/- 0.31</td><td style='padding:4px 8px;border-bottom:1px solid #334155;text-align:right'>-0.12</td><td style='padding:4px 8px;border-bottom:1px solid #334155;text-align:right'>+0.05</td><td style='padding:4px 8px;border-bottom:1px solid #334155;text-align:right'>+0.30</td></tr><tr><td style='padding:4px 8px;border-bottom:1px solid #334155'>rings</td><td style='padding:4px 8px;border-bottom:1px solid #334155;text-align:right'>86.56 +/- 0.47</td><td style='padding:4px 8px;border-bottom:1px solid #334155;text-align:right'>+0.18</td><td style='padding:4px 8px;border-bottom:1px solid #334155;text-align:right'>-0.06</td><td style='padding:4px 8px;border-bottom:1px solid #334155;text-align:right'>-0.10</td></tr><tr><td style='padding:4px 8px;border-bottom:1px solid #334155'>gauss-xor</td><td style='padding:4px 8px;border-bottom:1px solid #334155;text-align:right'>78.98 +/- 0.49</td><td style='padding:4px 8px;border-bottom:1px solid #334155;text-align:right'>+4.39</td><td style='padding:4px 8px;border-bottom:1px solid #334155;text-align:right'>+3.00</td><td style='padding:4px 8px;border-bottom:1px solid #334155;text-align:right'>-0.84</td></tr><tr><td style='padding:4px 8px;border-bottom:1px solid #334155'>digits-3v8</td><td style='padding:4px 8px;border-bottom:1px solid #334155;text-align:right'>97.55 +/- 0.54</td><td style='padding:4px 8px;border-bottom:1px solid #334155;text-align:right'>+0.56</td><td style='padding:4px 8px;border-bottom:1px solid #334155;text-align:right'>-0.41</td><td style='padding:4px 8px;border-bottom:1px solid #334155;text-align:right'>+0.12</td></tr></tbody>
  </table>

  <div style="margin-top:12px;padding:10px 12px;background:#442626;border-radius:8px;border-left:4px solid #ef4444;font-size:.87em">
    <b style="color:#fca5a5">Correction to v1 of this logbook.</b> v1 reported all 3 claims VERIFIED from a <b>single seed</b> on one easy moons dataset where every method scored an identical 99.75%. At 25 seeds across 4 datasets with paired tests and a random-push ablation, all three claims are <b>PARTIAL</b>. The mechanism is real and measurable; the advantage attributed to it is not.
  </div>
</div>