[AdaWorldAPI]

S3a–S3c: DR64³ Cartesian reference, holographic focus and optional DR256³ validation

Status: research plan only. No cubic ABI, new global ID, resident cube or performance result.

Typed geometry and source-backed constraints

Existing Boolean relcomp::compose already computes R:X→Y, S:Y→Z with Y contracted (a conceptual X×Y×Z iteration space). Proposed triangle (R∘S)∧Q requires explicit closing relation Q:X→Z and guards not automatically guaranteed by Chain adjacency. DR64² is 512 B, DR256² is 8 KiB; logical DR64³ would be 32 KiB, DR256³ 2 MiB, not mandated allocations. Quadratic ratio 1/16, full cube 1/64. The 64-thought occupancy plane is a distinct representation.

OP-ONNX-4 unresolved: relcomp.rs and dir-rmap-probe/src/compose.rs enumerate set-bit positions; qr.rs prohibits positional scanning under D-RU-6. Determine whether verified intra-register enumeration is admissible separately from forbidden value-steered cross-table gathers. No assumption or rule change without contract owner.

New compiled-palette identity contract

AdaWorldAPI/ndarray Burn has d_head-keyed CompiledAttention and shape-matched CompiledLinear (source inspected). DR global-to-local binding must not inherit these insufficient keys. Admit only witnessed model, layer, weight/codebook/calibration revision, assignment, source frame, focus/scope and typed semiring; changed source or stale codebook must refuse. Existing BGZ/Helix 256² metric-LUT proofs are not DR256³ or ONNX MatMul certification.

Gates: independent scalar Cartesian oracle; legal address/focus dependency closure; missing-coordinate and mismatched-identity negative cases; explicitly measured binding/expansion/materialization. DR256³ utility separately falsifiable and not required for #1748 or #1749.

Related #1744 / #1745 / #1750 S0 conflict, #1746 semantics, #1748 physical shootout, #1749 identity reuse, EXO-50. Docs only, no authorization.

---

Dependency clarification: #1751

The first Optimum → BGZ/CompiledLinear one-layer probe does not require DR64³, DR256³ or another global identity representation. DR64/DR256 focus is an optional later physical execution strategy requiring its own source/focus/scope/identity witnesses. No palette-byte-to-foreign-ordinal reinterpretation. The source-matched compiler sidecar cannot bypass ADMISSION B.

⊘ Supersession — D-NFC (2026-10-23, operator-approved architecture correction)

Canonical dependency diagram and root assumptions: #1753 (d-native-first-dependency-correction-v1.md); shipped / lab / missing seams: d-native-repr-contract-v1.md there.

P0 invariants. (1) ONNX is an external mathematical compatibility specification; the existing R2IL polyfill is the native starting point; no ONNX graph, Tract runtime or Optimum dependency in the internal execution path. (2) Quack binding + DR64³ + Palette256³ is the intended native representation; existing R2IL, mask-risc, BGZ, Helix and ndarray primitives are reused; implemented components and unimplemented integration contracts are kept apart.

[1] R2IL router (once per ClassId; strategies SHIPPED, #1752) → [2] Quack binding (SHIPPED; store MISSING)
→ [3] DR64³ geometry (LAB; 3rd axis PROPOSED) → [4] Palette256³ (SHIPPED; DR binding MISSING)
→ [5] native execution: mask-risc + ndarray::simd (SHIPPED) → [6] egglog laws (LAB, Boolean only)
→ [7] ONNX conformance adapters (PROPOSED) → [8] Tract/ORT oracles, Optimum (OPTIONAL)

⊘ Layer [1] correction (2026-10-23, #1753 commit ace69d64c, #1752 plan §10). Layer [1] is the R2IL router, which resolves a route once per ClassId: DIRECT or COMPOSE onto ndarray::simd, LOWER to r2sleigh only for machine semantics, or REFUSE. lance-graph-r2il procedures + FoldDialect are one strategy beneath it, not the layer. No operation in this plan needs R2Sleigh.

For this PR:

1. DR64³ is the intended native relational / focus representation (layer [3] of the canonical diagram), not an optional ONNX-side optimization (A5). Only its physical realization (dense, sparse, tiled) is a measured choice.
2. Its architectural parents are the native polyfill (lab(mpf): D-MPF-1 — the Stage 1 math polyfill already exists; one expression, oracle, refusals, cost split #1752) and the Quack binding (D-NFR S1–S3), not ONNX intake (docs(onnx) S1–S2: semantic intake and minimal RelAlg polyfill plan #1746), which is a downstream conformance adapter.
3. What exists today (D-NFR §2): frame seal, focus and scope guards, typed method axes and the 64×64 register are LAB (dr-math-binding-probe); the 256-slot global → local focus is LAB (dir-rmap-probe); the third axis has no type (the register is DR64²). Nothing here claims more.
4. Identity witnesses are native: a sealed frame plus a codebook generation (the latter MISSING, D-NFR S15), not ONNX model / layer / node identity (A7). ADMISSION B is unchanged: a DR ordinal is never a palette index.

Where this description or the plan says otherwise, it is superseded at the marked ⊘ points (plan commit 7939ebcc8). Nothing is merged, no ABI or execution contract changes, and the #1744 / #1745 / #1750 S0 dispute is unchanged.

[coderabbitai]

Important

• 🔍 Trigger review

This repository does not receive automatic reviews because it has fewer than 10 stars.

⚙️ Run configuration

• Source mode: Auto
• Repository: Unavailable
• Branch: Unavailable
• File: No configuration file from this repository applied
• Commit: Unavailable
• Layers: Organization UI
• Review profile: CHILL
• Plan: Advanced
• Run ID: 51329d58-9739-42d5-8c34-e511d929d7d9

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Warning

Billing warning: we have not been able to collect payment for this subscription for more than 72 hours. Please update the payment method or pay any pending invoices in Billing to avoid service interruption.

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.}

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-10-23T22:13:43.255166Z	7939ebc	Draft marked ready

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.