feat(site): field catalog + Phase-1 comparison pages by jphein · Pull Request #259 · techempower-org/multipass-structural-memory-eval

jphein · 2026-07-04T15:22:02Z

Publishes the two comparison pages to the Pages site: /site/catalog.html (ranked field catalog — next systems to compare, metric-type badges, engram-2 deaccession record) and /site/phase1.html (OMEGA vs mempalace vs Go server structural three-way, basis-attributed). Cross-linked with the field explorer.

🤖 Generated with Claude Code

…pages The ranked memory-systems catalog (with the engram-2 deaccession card and metric-type badges) and the OMEGA vs mempalace vs Go-server structural comparison, previously served locally. Self-contained pages alongside the field explorer. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

Copilot

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

coderabbitai · 2026-07-04T15:22:10Z

Warning

Review limit reached

@jphein, you've reached your PR review limit, so we couldn't start this review.

Next review available in: 27 minutes

Enable usage-based reviews in Billing to review now. Otherwise, wait until the next included review is available.
You're only billed for reviews past your plan's rate limits ($0.25/file).

How can I continue?

After more reviews become available, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews.

How do review limits work?

CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability.

For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window.

Please refer docs for additional details.

Review details

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro Plus

Run ID: deab679d-a899-49c5-a6fe-bf25ee628a57

📥 Commits

Reviewing files that changed from the base of the PR and between a15dd17 and bc710ab.

📒 Files selected for processing (2)

docs/site/catalog.html
docs/site/phase1.html

✨ Finishing Touches

🧪 Generate unit tests (beta)

Create PR with unit tests
Commit unit tests in branch feat/site-catalog-phase1

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

_{Comment @coderabbitai help to get the list of available commands.}

gemini-code-assist

Code Review

This pull request adds two new documentation pages, catalog.html and phase1.html, which present a field catalog of AI memory systems and a structural comparison between OMEGA, mempalace, and a Go server. The feedback focuses on improving the HTML and CSS structure of both files. Specifically, both documents are missing standard HTML boilerplate (such as DOCTYPE, html, head, and body tags), which can trigger Quirks Mode in browsers. Additionally, the feedback recommends using semantic table elements like thead and tbody for better accessibility, and moving CSS @Keyframes definitions outside of @media blocks to ensure broader browser compatibility.

Important

The consumer version of Gemini Code Assist on GitHub is being sunset. Starting June 18, 2026, new organization installations will be blocked, and all code review activity will officially cease on July 17, 2026.
For more details on the timeline and next steps, please review the Help Documentation.

gemini-code-assist · 2026-07-04T15:23:48Z

+<title>Field Catalog — AI Memory Systems</title>
+<style>


The HTML document is missing the standard document structure boilerplate, including the <!DOCTYPE html> declaration, <html> tag (with a lang attribute for accessibility), and <head> tag. Without <!DOCTYPE html>, browsers will render the page in Quirks Mode instead of Standards Mode, which can cause layout inconsistencies and rendering bugs across different browsers. Additionally, adding a <meta name="viewport"> tag is essential for proper mobile responsiveness.

<!DOCTYPE html> <html lang="en"> <head> <meta charset="UTF-8"> <meta name="viewport" content="width=device-width, initial-scale=1.0"> <title>Field Catalog — AI Memory Systems</title> <style>

gemini-code-assist · 2026-07-04T15:23:48Z

+</style>
+
+<div class="wrap">


Close the <head> tag and open the <body> tag to establish a standard HTML document structure.

</style> </head> <body> <div class="wrap">

gemini-code-assist · 2026-07-04T15:23:48Z

+  <p><strong>Provenance.</strong> Claims and statuses verified 2026-07-03 against <a href="https://github.com/techempower-org/mempalace/blob/main/docs/research/2026-05-24-memory-system-benchmarks.md">our benchmark-methodology research</a>, <a href="https://github.com/techempower-org/mempalace/blob/main/docs/ECOSYSTEM.md">ECOSYSTEM.md</a>, the <a href="https://github.com/techempower-org/multipass-structural-memory-eval/tree/main/docs">multipass eval findings</a>, and live GitHub/Wayback/crates/npm archaeology performed today.</p>
+  <p>Metric badges follow the retrieval-vs-QA distinction argued in MemPalace <a href="https://github.com/MemPalace/mempalace/discussions/747">discussion #747</a>. Systems marked <em>self-reported</em> have published numbers we could not independently confirm.</p>
+</footer>
+</div>


Close the <body> and <html> tags at the end of the document.

</div> </body> </html>

gemini-code-assist · 2026-07-04T15:23:48Z

+<title>Phase 1 — OMEGA vs mempalace vs Go server</title>
+<style>


The HTML document is missing the standard document structure boilerplate, including the <!DOCTYPE html> declaration, <html> tag (with a lang attribute for accessibility), and <head> tag. Without <!DOCTYPE html>, browsers will render the page in Quirks Mode instead of Standards Mode, which can cause layout inconsistencies and rendering bugs across different browsers. Additionally, adding a <meta name="viewport"> tag is essential for proper mobile responsiveness.

<!DOCTYPE html> <html lang="en"> <head> <meta charset="UTF-8"> <meta name="viewport" content="width=device-width, initial-scale=1.0"> <title>Phase 1 — OMEGA vs mempalace vs Go server</title> <style>

gemini-code-assist · 2026-07-04T15:23:48Z

+</style>
+<div class="wrap">


Close the <head> tag and open the <body> tag to establish a standard HTML document structure.

</style> </head> <body> <div class="wrap">

gemini-code-assist · 2026-07-04T15:23:48Z

+  <p><strong>Provenance.</strong> All JSONs in the session scratch (<span class="mono">phase1/</span>): omega × 6 (incl. emergent report + ground-truth Cat 3), mempalace real-KG × 5, go-server × 5. Runners: <span class="mono">run_omega_good_dog_structural.py</span> (existing) and its new twin <span class="mono">run_mempalace_server_good_dog_structural.py</span> (same rows, same scorer). OMEGA 1.4.15 · palace-daemon @ familiar (read-only) · Go server @ localhost:8000 (disposable instance, wiped).</p>
+  <p>Phase 2 (pending approval): Cat 9 + LongMemEval QA with reader gpt-4o-mini / judge gpt-4.1-mini — the actual hostile rerun of OMEGA's 95.4%.</p>
+</footer>
+</div>


Close the <body> and <html> tags at the end of the document.

</div> </body> </html>

gemini-code-assist · 2026-07-04T15:23:48Z

+  @media (prefers-reduced-motion: no-preference){
+    .gone::after{animation:stamp .38s cubic-bezier(.2,1.6,.4,1) .45s backwards}
+    @keyframes stamp{from{transform:rotate(24deg) scale(2.4); opacity:0} to{transform:rotate(24deg) scale(1); opacity:.82}}
+  }


Defining @keyframes inside a @media block is valid in modern CSS, but historically it has compatibility issues in older browsers (e.g., older Safari/Chrome/Firefox versions). To ensure maximum compatibility and clean separation of concerns, it is a best practice to define @keyframes at the root level of the stylesheet and only reference the animation name inside the media query.

Suggested change

@media (prefers-reduced-motion: no-preference){

.gone::after{animation:stamp .38s cubic-bezier(.2,1.6,.4,1) .45s backwards}

@keyframes stamp{from{transform:rotate(24deg) scale(2.4); opacity:0} to{transform:rotate(24deg) scale(1); opacity:.82}}

}

@keyframes stamp{from{transform:rotate(24deg) scale(2.4); opacity:0} to{transform:rotate(24deg) scale(1); opacity:.82}}

@media (prefers-reduced-motion: no-preference){

.gone::after{animation:stamp .38s cubic-bezier(.2,1.6,.4,1) .45s backwards}

}

gemini-code-assist · 2026-07-04T15:23:48Z

+    <table class="ledger">
+      <tr><td>Claimed</td><td class="mono">LongMemEval-S R@5 <strong>0.990</strong> · R@10 0.998 · MRR 0.946 — config: Gemini&nbsp;Embed&nbsp;2 + FTS5 + RRF (not the pure-local mode) · LoCoMo-QA 74.5% (strict judge, 200q subset)</td></tr>
+      <tr><td>Entered orbit</td><td>Posted by <span class="mono">@longevityboris</span> in MemPalace <a href="https://github.com/MemPalace/mempalace/issues/29">issue #29</a>, Apr 11 — the benchmark-methodology wars</td></tr>
+      <tr><td>Last seen</td><td>Apr 12, 2026 (dial481 browsing its <span class="mono">feat/mlx-reranker</span> branch)</td></tr>
+      <tr><td>Gone by</td><td>Jun 10, 2026 — 404 recorded in our commit <span class="mono">9781ce5</span>; zero Wayback snapshots; no surviving forks, crate, or package</td></tr>
+      <tr><td>Survives in</td><td>Our palace’s verbatim capture of its README claims, and these notes</td></tr>
+    </table>


Wrap the table rows in a <tbody> element to ensure semantic correctness and better browser parsing.

<table class="ledger"> <tbody> <tr><td>Claimed</td><td class="mono">LongMemEval-S R@5 <strong>0.990</strong> · R@10 0.998 · MRR 0.946 — config: Gemini Embed 2 + FTS5 + RRF (not the pure-local mode) · LoCoMo-QA 74.5% (strict judge, 200q subset)</td></tr> <tr><td>Entered orbit</td><td>Posted by <span class="mono">@longevityboris</span> in MemPalace <a href="https://github.com/MemPalace/mempalace/issues/29">issue #29</a>, Apr 11 — the benchmark-methodology wars</td></tr> <tr><td>Last seen</td><td>Apr 12, 2026 (dial481 browsing its <span class="mono">feat/mlx-reranker</span> branch)</td></tr> <tr><td>Gone by</td><td>Jun 10, 2026 — 404 recorded in our commit <span class="mono">9781ce5</span>; zero Wayback snapshots; no surviving forks, crate, or package</td></tr> <tr><td>Survives in</td><td>Our palace’s verbatim capture of its README claims, and these notes</td></tr> </tbody> </table>

gemini-code-assist · 2026-07-04T15:23:48Z

+<div class="tablebox"><table>
+  <tr><th>Reading</th><th>OMEGA<br><span style="text-transform:none">good-dog emergent</span></th><th>mempalace<br><span style="text-transform:none">real KG (sample)</span></th><th>Go server<br><span style="text-transform:none">basis=taxonomy (no auto-extraction)</span></th></tr>
+  <tr><td>Graph size (entities / edges)</td><td class="num">18 / 54</td><td class="num">12,659 / 10,000*</td><td class="num na">3 / 1 †</td></tr>
+  <tr><td>Cat 4 — canonical collisions</td><td class="num good">0</td><td class="num bad">222</td><td class="num na">n/m †</td></tr>
+  <tr><td>Cat 4 — required-field coverage</td><td class="num good">1.00</td><td class="num good">1.00</td><td class="num">1.00 †</td></tr>
+  <tr><td>Cat 4 — edge-type entropy (bits)</td><td class="num">1.24</td><td class="num good">3.43</td><td class="num na">0.0 †</td></tr>
+  <tr><td>Cat 5 — components / largest</td><td class="num good">1 / 18</td><td class="num bad">6,072 / 3,900*</td><td class="num na">2 / 2 †</td></tr>
+  <tr><td>Cat 3 — structural contradiction detection</td><td class="num good">1.00 <span class="pill p-info">Δ+1.00 vs flat</span></td><td class="num na">0 pairs — no contradicts edge type in prod KG</td><td class="num na">n/m †</td></tr>
+  <tr><td>Cat 3 — vs good-dog ground truth (theme recall)</td><td class="num">0.50 <span class="pill p-warn">caught DCM, missed dominance</span></td><td class="num na">different basis</td><td class="num na">emergent KG empty — no extraction</td></tr>
+  <tr><td>Cat 6 — supersession completeness</td><td class="num na">0 seeded — corpus text carries none emergently</td><td class="num na">no supersedes edge type</td><td class="num na">n/m †</td></tr>
+</table></div>


Add <thead> and <tbody> elements to the table to ensure semantic correctness, better accessibility for screen readers, and consistent rendering.

<div class="tablebox"><table> <thead> <tr><th>Reading</th><th>OMEGA<br><span style="text-transform:none">good-dog emergent</span></th><th>mempalace<br><span style="text-transform:none">real KG (sample)</span></th><th>Go server<br><span style="text-transform:none">basis=taxonomy (no auto-extraction)</span></th></tr> </thead> <tbody> <tr><td>Graph size (entities / edges)</td><td class="num">18 / 54</td><td class="num">12,659 / 10,000*</td><td class="num na">3 / 1 †</td></tr> <tr><td>Cat 4 — canonical collisions</td><td class="num good">0</td><td class="num bad">222</td><td class="num na">n/m †</td></tr> <tr><td>Cat 4 — required-field coverage</td><td class="num good">1.00</td><td class="num good">1.00</td><td class="num">1.00 †</td></tr> <tr><td>Cat 4 — edge-type entropy (bits)</td><td class="num">1.24</td><td class="num good">3.43</td><td class="num na">0.0 †</td></tr> <tr><td>Cat 5 — components / largest</td><td class="num good">1 / 18</td><td class="num bad">6,072 / 3,900*</td><td class="num na">2 / 2 †</td></tr> <tr><td>Cat 3 — structural contradiction detection</td><td class="num good">1.00 <span class="pill p-info">Δ+1.00 vs flat</span></td><td class="num na">0 pairs — no contradicts edge type in prod KG</td><td class="num na">n/m †</td></tr> <tr><td>Cat 3 — vs good-dog ground truth (theme recall)</td><td class="num">0.50 <span class="pill p-warn">caught DCM, missed dominance</span></td><td class="num na">different basis</td><td class="num na">emergent KG empty — no extraction</td></tr> <tr><td>Cat 6 — supersession completeness</td><td class="num na">0 seeded — corpus text carries none emergently</td><td class="num na">no supersedes edge type</td><td class="num na">n/m †</td></tr> </tbody> </table></div>

Copilot AI review requested due to automatic review settings July 4, 2026 15:22

Copilot AI reviewed Jul 4, 2026

fix(site): catalog bench strip → link Phase-1 results

bc710ab

gemini-code-assist Bot reviewed Jul 4, 2026

View reviewed changes

jphein merged commit 30d7fcd into main Jul 4, 2026
4 checks passed

jphein deleted the feat/site-catalog-phase1 branch July 4, 2026 15:23

		<title>Phase 1 — OMEGA vs mempalace vs Go server</title>
		<style>

Uh oh!

Conversation

jphein commented Jul 4, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Uh oh!

coderabbitai Bot commented Jul 4, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Review limit reached

Uh oh!

gemini-code-assist Bot left a comment

Choose a reason for hiding this comment

Code Review

Uh oh!

gemini-code-assist Bot Jul 4, 2026

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist Bot Jul 4, 2026

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist Bot Jul 4, 2026

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist Bot Jul 4, 2026

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist Bot Jul 4, 2026

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist Bot Jul 4, 2026

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist Bot Jul 4, 2026

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist Bot Jul 4, 2026

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist Bot Jul 4, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

coderabbitai Bot commented Jul 4, 2026 •

edited

Loading