docs(knowledge): probe M4 result — FAIL DIRECTION (terminates at TWIG, not HEEL)

[AdaWorldAPI]

Records probe M4 in .claude/knowledge/bf16-hhtl-terrain.md per that file's own update protocol. One file, one commit — no code.

Why now

M4 ("HHTL termination: what % at each level?") has sat NOT RUN in the probe queue while its own process rule required any bucketing-strategy change to run it first or be labelled CONJECTURE. Its harness — openstreetmap-website-rs/src/bin/tier_probe.rs — was written for this entry and never executed.

It was run because a downstream consumer (q2's cockpit /api/osm/features) changed its decimation strategy and the rule applied. Ran on two real extracts, deliberately one dense and one sparse.

Result

Features per tile, by cascade tier:

tier	Berlin — city, 2.52 M features	Iceland — overland, 0.65 M features
	tiles / med / p95 / max / fit≤30	tiles / med / p95 / max / fit≤30
heel z8	2 / 1,564,647 / — / 1,564,647 / 0.0 %	58 / 3,838 / 34,985 / 202,296 / 20.7 %
hip z16	8,065 / 206 / 996 / 3,844 / 16.4 %	178,962 / 1 / 8 / 1,067 / 95.2 %
twig z24	2,435,641 / 1 / 1 / 20 / 99.7 %	649,093 / 1 / 1 / 7 / 99.9 %
leaf z32	2,513,559 / 1 / 1 / 11 / 99.8 %	652,314 / 1 / 1 / 7 / 99.9 %

Against the stated gate (>60% HEEL pass, >60% LEAF fail) this is the fail direction: 99.7 % of Berlin's twig cells hold exactly one feature.

Two findings carried forward

1. There is no coarse bucket to compress into below hip. Occupancy jumps ~200× from twig (1) to hip (206), then ~7,600× to heel. Any LOD or decimation scheme therefore has exactly one useful bucketing level, and "bucketing > resolution" (Iteration 5 in that doc) holds only at and above the hip cell.
2. Density is a property of the extract, not of the zoom. Berlin and Iceland differ ~200× at hip and converge to 1 by twig, so a decimation rule keyed on zoom alone is mis-specified for one of them; cell occupancy is the discriminating quantity.

Scope, stated so it is not over-read

This is M4 in its OSM point-feature form. It says where a geographic cascade terminates on real planet data. It says nothing about HHTL termination for embedding fingerprints, which is what P2–P4 address and which remains NOT RUN. The queue row is marked RAN 2026-08-11 -> FAIL DIRECTION rather than PASS/FAIL, because the gate's own phrasing assumes the embedding form.

Consumer

Recorded in the entry: q2's cockpit-server::osm_features::row_budget, whose overview branch is now labelled CONJECTURE against finding (1) — it strides rows where the cascade answer is one representative per occupied hip cell. That change merged as AdaWorldAPI/q2#110.

Append-only: the queue row is updated in place per the update protocol (Record result in the Status column), and the detail lands as a new section beneath it. No prior entry is edited or removed.

🤖 Generated with Claude Code

https://claude.ai/code/session_01NMeiLmtDKhomJNSo2ecbJw

---

Generated by Claude Code

Summary by CodeRabbit

• Documentation
  • Recorded Probe M4 results from August 11, 2026 using OpenStreetMap data.
  • Documented that the HEEL termination gate failed, with termination occurring primarily at TWIG.
  • Added regional findings for Berlin and Iceland, including 99.7–99.9% single-feature TWIG cells.
  • Clarified that HIP is the only useful lower-resolution bucketing level.
  • Noted that occupancy depends on extract density, not zoom alone.
  • Limited the findings to OpenStreetMap point features, excluding embedding fingerprints.

[coderabbitai]

Caution

Review failed

The pull request is closed.

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 3a9fff51-f760-47ef-9cca-7da3c8fba9c9

📥 Commits

Reviewing files that changed from the base of the PR and between b1bcee4 and 7bc2089.

📒 Files selected for processing (1)

• .claude/knowledge/bf16-hhtl-terrain.md

---

📝 Walkthrough

Walkthrough

Probe M4 is now recorded as run on 2026-08-11. The results show that OSM point-feature cells terminate primarily at TWIG, not HEEL, in Berlin and Iceland. The document records occupancy, bucketing, density, scope, and row_budget findings.

Changes

Terrain probe findings

Layer / File(s)	Summary
Record Probe M4 results .claude/knowledge/bf16-hhtl-terrain.md	Records the OSM probe methodology, Berlin and Iceland measurements, TWIG termination results, HIP bucketing behavior, density relationship, scope limitation, and affected row_budget consumer.

Estimated code review effort: 1 (Trivial) | ~3 minutes

Suggested reviewers: claude

Poem

A rabbit records the terrain with care,
TWIG cells gather nearly everywhere.
HIP marks the useful chart,
Density guides the crowded part.
OSM findings now rest there.

---

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

• X
• Mastodon
• Reddit
• LinkedIn

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

[cursor]

Bugbot couldn't run - usage limit reached

Bugbot is counted against Cursor usage for this user or team, and this run hit a usage or spend limit.

A user or team admin can review and increase usage limits in the Cursor dashboard.

(requestId: serverGenReqId_c18f3031-2f9b-4f17-ab32-fd87a0f896e8)