The particle/wave click → ClassView::compute_dag (the one Core gap) + the electricity-cascade join#539
Merged
Merged
Conversation
…ick (8-agent 5+3)
Verdict (unanimous; cascade-architect grounded in shipped ndarray code): the
"particle/wave click" is two real mechanisms + one costume.
COLLAPSES (legitimate, at the ADDRESS only):
- Address ROUTER [H]: one shift/mask prefix-distance over NodeGuid serves
2-axis spatial (OSM x/y) AND semantic (PQ-pair); semantic GREEN-shipped
(cam_pq DistanceTables). D-BOTHCASC real at the key.
- Field ENCODER [H]: WHT/bipolar-phase pyramid synthesizes field from
address-phase (never stored) + magnitude tenant; two-algebra Markov-fenced.
MUST STAY DISTINCT (allocation, not conservation): ValueSchema Cognitive vs
Compressed; EdgeBlock CoarseOnly vs PQ; ActionInvocation (RBAC+MUL Rubicon) vs
deterministic recurrence. Different byte ranges, opposite sides of the firewall.
STRIP (rhyme): "particle/wave duality" (no norm-preserving T(Class)<->field);
"simultaneously" (classid dispatch, one domain per node); "Heisenberg" (N<=sqrt(d)/4
is a capacity floor, not dx*dp -> demote to [S]).
TWO CRACKS for the named 3D consumer (OSM->gaussian-splat->electricity):
shipped spatial is 3D Hilbert (hilbert.rs N=3), NOT the 2-axis 256x256 tile ->
it's a RELATED-BUT-DISTINCT 3-axis sibling cascade; and spatial distance is
SSE/cert-gated, not LUT yet.
HARD-PROOF: the bit-partition identity (centroid 4-ary tree == WHT sign-pyramid
bit-grouping over the same tier-byte; fix "nibbles"->"2-bit crumbs"). Plus
PROBE-CODEBOOK-44, pin-spatial-to-2-axis, norm-preserving-T (predicted fail),
disjoint-schema regression, revert ReadMode Full POC. Core: ClassView::
{axis_binding,centroid_codebook}, a Field ValueSchema preset, the Jirak reword.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CcpLeEC3XK8Eye53GKBVvi
…e proof
A spreadsheet projected as a shader maps every piece of the stack onto a
shipped/named primitive (mechanism, not rhyme): cell(row,col)=2-axis router
node; value=particle tenant; formula==ActionInvocation (DO arm); dependency
DAG=ClassView::compute_dag (THE named Core gap); recompute=cycle-aware write +
field-wave; the projected sheet=CognitiveShader dispatch.
Ideal first proof because: (1) 2-axis = the router's PROVEN home (sidesteps the
3D-Hilbert crack from E-OGAR-ROUTER-ENCODER; OSM->splat->electricity is the 3D
sibling cascade); (2) formulas ARE compute_dag -> lands the one Core gap on the
cleanest case; (3) universal -- Odoo @api.depends / medcare lab-trends / woa
calc / q2 cells all reduce to a sheet (E-AR-DO-WIRING consumers collapse to it);
(4) recompute exercises write_row end-to-end.
Scope line: formula SEMANTICS are general compute via DO arm/UnifiedStep, NOT
the Walsh-Hadamard field -- "shader projection" = dependency-driven recompute
dispatch, not field synthesis (never the physics costume E-OGAR-ROUTER-ENCODER
stripped).
Probe PROBE-EXCEL-COMPUTE-DAG: minimal sheet (cells=SoA rows, formulas=ActionDef
+depends_on, compute_dag topological recompute gated by write_row(cycle));
success = edit dirties+recomputes dependents in cycle-stamped topological order,
loop rejected at registry-build, Cognitive/Compressed disjoint. Lands
ClassView::compute_dag for all computed-field AR consumers.
Weaves in prior-iteration details: E-OGAR-ROUTER-ENCODER, E-AR-DO-WIRING,
E-SOA-CYCLE-OWNERSHIP, the converged deliverables (ClassView::{axis_binding,
centroid_codebook}, Field ValueSchema preset, Jirak reword), lite-unified.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CcpLeEC3XK8Eye53GKBVvi
E-CHESS-TENSOR-PROVEN: chess AI is the proven-at-scale instance of router(8x8 grid) + field(feature planes) + incremental-recompute(NNUE) + search(MCTS). The 4 species = the session's 4 facets: bitboards=particle register, AlphaZero/Leela CNN=field/wave (shader projection), Stockfish NNUE= the cycle-aware incremental recompute (compute_dag), transformer=EdgeBlock/ attention coupling (bgz-tensor). NNUE's "only small input changes between neighboring positions, incrementally update" IS PROBE-EXCEL-COMPUTE-DAG proven at world-champion strength. 2-axis (rank x file) -> validates the architecture without touching the open 3D-Hilbert/bit-partition questions. Excel stays the consumer demonstrator; chess is the validator + NNUE-incremental design ref. probe-excel-compute-dag-v1: the named first proof for ClassView::compute_dag. Inc0 compute_dag(classid) topological manifest from depends_on+emitted_by (registry rejects cycles, layout-preserving default-method); Inc1 sheet harness (cells=SoA rows 2-axis, formulas=ActionDef+depends_on); Inc2 recompute-on-edit gated by write_row(cycle); Inc3 success = cycle-stamped topological dirty- propagation + loop-rejected-at-build + Cognitive/Compressed disjoint + incremental==full parity (the NNUE invariant). Lands the gap for every computed-field AR consumer at once. Board: INTEGRATION_PLANS prepended. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CcpLeEC3XK8Eye53GKBVvi
…probe Inc 0)
Inc 0 of probe-excel-compute-dag-v1 — lands the one named Core gap
(ClassView::compute_dag), the recompute-dispatch home every computed-field
AR consumer needs (Odoo @api.depends / medcare / woa / q2, all reduce to a
sheet) and the NNUE-incremental existence-proof shape.
- ComputeEdge {target: u8, inputs: &'static [u8]}: harvest-sourced recompute
edge (emitted_by target <- depends_on inputs), const-constructible like
MethodSig/ActionDef. Field positions index the class FieldMask.
- ClassView::compute_dag(class) -> &[ComputeEdge]: default &[] (zero-fallback);
layout-preserving default-method (mirrors value_schema), stores nothing on
the row, zero NODE_ROW_STRIDE/ENVELOPE_LAYOUT_VERSION impact.
- compute_dag_is_acyclic: the registry-build gate -- a cyclic recompute DAG
(formula loop / @api.depends cycle / self-loop) is rejected at build (Kahn
over <=64 positions, allocation-free; out-of-range ignored, no panic).
+4 tests (default-empty, acyclic-chain, 2/self/3-cycle rejected, out-of-range
ignored); 10/10 class_view; clippy -D warnings + fmt clean. Board:
LATEST_STATE Contract Inventory prepended.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CcpLeEC3XK8Eye53GKBVvi
…R, not just the acyclic gate Inc 0 landed `compute_dag_is_acyclic` (the registry-build cycle gate). The sheet/cascade harness (probe-excel-compute-dag-v1 Inc 2) needs the actual topological recompute ORDER: edit a cell → recompute its transitive dependents in an order where each target follows everything it depends on. `compute_dag_topo_order(edges) -> Option<Vec<u8>>` is Kahn over the same 64-bit field masks (allocation = one Vec of the ≤64 target positions). Returns the target positions in a valid recompute order; `None` on a cycle (mirroring `compute_dag_is_acyclic == false`). Leaves (positions only ever read, never a target) are excluded — they are the already-present values a recompute reads. This is the same structure Stockfish NNUE proves at world-champion strength: "only small input changes between neighboring positions → incrementally update the affected accumulator dependents in dependency order." Re-derivation, not invention. Tests: chain (f0→f1→f2, f1 before f2), diamond (both precedents before the join), empty (Some(vec![]) not None), cycle (None). 13/13 class_view green, clippy clean. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CcpLeEC3XK8Eye53GKBVvi
…— same topological recompute, plus the Weyl bound that certifies incrementality The crates/perturbation-sim outage simulator is the physical instance of the same dependency-driven topological recompute that ClassView::compute_dag + compute_dag_topo_order + write_row express abstractly: simulate_outage round loop ≡ compute_dag_topo_order dispatch trip_round[e] (which round) ≡ topological generation seed trip = rank-1 E on L ≡ the dirty seed write_row(seed, cycle) spectral_perturbation Weyl ≡ the NNUE incremental ≡ full bound PerturbationShape.node_field ≡ the wave/field readoff splat::morton2 ≡ the 2-axis router address sketch::fwht + walsh pyramid ≡ the deterministic field encoder witness::particle_equals_wave≡ the particle/wave click, on a real field The honest framing: the two crates are complementary halves. perturbation-sim does the EXACT FULL recompute each round (robust, no LODF drift) but ships the certification apparatus — its Weyl/Davis-Kahan bounds are exactly what an incremental scheme needs to prove it equals the full recompute. compute_dag is the incremental dispatch; the Weyl bound certifies why it's sound. Stockfish NNUE proves the incremental side at scale. No dependency added: perturbation-sim stays zero-dep / workspace-excluded. This is a doc-level join (COMPUTE_DAG_MAPPING.md) + epiphany, not a contract import. The crate already self-guards every scope line (no measured speed claim in witness.rs; numeric witness arc explicitly NOT the contract WitnessTable per I-VSA-IDENTITIES; Jirak not IID Berry-Esseen per I-NOISE-FLOOR-JIRAK). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CcpLeEC3XK8Eye53GKBVvi
|
Warning Review limit reached
More reviews will be available in 58 minutes and 15 seconds. Learn how PR review limits work. Your organization has used up its prepaid credits, and credit purchases are no longer available. Enable the review add-on in the billing tab to keep reviews running — you're only billed for reviews past your plan's rate limits ($0.25/file). ⌛ How to resolve this issue?After more reviews become available, a review can be triggered using the 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 credits. 🚦 How do rate limits work?CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan refill rate. 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, the refill rate gradually slows as usage increases. The highest same-day bursts are limited more strictly. Please see our Fair Usage Limits Policy for further information. ℹ️ Review info⚙️ Run configurationConfiguration used: Organization UI Review profile: CHILL Plan: Pro Plus Run ID: 📒 Files selected for processing (6)
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. Comment |
…ng frame the click epiphanies reference E-AR-DO-WIRING (ruff static + ARM dynamic meet at SPO; AR Class splits along DOLCE into THINK+DO; consumers land off the harvest) was stranded on a stale branch (claude/ar-do-wiring-epiphany) cut from a pre-#538 main — its diff against current main showed #538's files as deletions. The only real content was this single docs-only epiphany (+53 on EPIPHANIES.md), which the particle/wave click epiphanies in this PR cross-reference 5 times. Placed after E-OGAR-ROUTER-ENCODER (chronological: it is the prior-art frame the router/encoder + Excel + chess + perturbation entries build on). Makes this PR's arc self-contained — no dangling refs to an abandoned branch. The stale branch can be deleted. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CcpLeEC3XK8Eye53GKBVvi
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What this lands
Hardens the "particle/wave click" into a buildable Core extension, then proves it has a real physical instance.
Epiphanies (docs/board, prepend-only):
E-OGAR-ROUTER-ENCODER— 8-agent 5+3 verdict: the click is a domain-agnostic address router (shift/mask prefix-distance, 6 bytes = CAM-PQ 6×256 code) + a deterministic field encoder (bipolar-phase Walsh–Hadamard pyramid; sign = XOR =vsa_bind, magnitude =vsa_bundle). The physics duality framing is a costume → stripped. The click is REAL and mostly already in code.E-EXCEL-SHADER-PROJECTION— a spreadsheet projected as a shader is the clean 2-axis existence proof for the router + DO arm +compute_dag+ cycle-aware write. Sidesteps the open 3D-Hilbert / bit-partition questions.E-CHESS-TENSOR-PROVEN— chess AI is the proven-at-scale instance: bitboards = particle register, AlphaZero/Leela CNN = field/wave, Stockfish NNUE = the incremental cycle-aware recompute (compute_dag) proven at world-champion strength, transformer =EdgeBlock/attention.E-PERTURBATION-CASCADE-IS-COMPUTE-DAG— thecrates/perturbation-simoutage simulator IScompute_dagrunning on a real physical field, and it ships the Weyl bound that certifies the incremental recompute.Code (the one named Core gap,
lance-graph-contract::class_view, all additive / layout-preserving):ComputeEdge { target: u8, inputs: &'static [u8] }— the const-constructible recompute manifest.compute_dag_is_acyclic(edges) -> bool— Kahn over 64-bit field masks, allocation-free; the registry-build cycle gate.compute_dag_topo_order(edges) -> Option<Vec<u8>>— the recompute ORDER (the harness needs the order, not just the bool);Noneon a cycle; leaves excluded.ClassView::compute_dag(class) -> &[ComputeEdge]default method (zero-fallback empty manifest), mirroringfields/value_schema.Plan:
probe-excel-compute-dag-v1— the named first proof forcompute_dagon the 2-axis sheet.Join (no dependency):
crates/perturbation-sim/COMPUTE_DAG_MAPPING.md—simulate_outageround loop ≡compute_dag_topo_orderdispatch;trip_round≡ topo generation; seed trip (rank-1EonL) ≡ the dirtywrite_row;spectral_perturbationWeyl/Davis–Kahan ≡ the NNUE incremental≡full bound;node_field≡ the wave;splat::morton2≡ the router;sketch::fwht≡ the field encoder;witness::particle_equals_wave≡ the click on a real field. perturbation-sim stays zero-dep / workspace-excluded — doc-level join only.Honesty / scope guards
witness.rsalready states this; inherited).WitnessTable(I-VSA-IDENTITIESregister-loss guard, already inwitness.rs).UnifiedStep— the recompute structure transfers, the formula content does not.I-NOISE-FLOOR-JIRAK).⚠ Must revert before merge (documented + test-guarded, not silent drift)
ClassView::value_schemadefault returnsValueSchema::Full— a TEMPORARY POC default (2026-06-15) so downstream consumers (odoo-rs / openproject-nexgen-rs / woa-rs / medcare-rs / q2) can transcode against a fully-populatedNodeRowduring this consumer-landing phase. The type-levelValueSchema::default()staysBootstrap, so substrate zero-fallback semantics are untouched — only the class→schema resolution default isFull. Revert this default toBootstrapand thevalue_schema_default_is_full_temporary_poctest together when the POC phase ends.Tests
13/13
class_view(3 new topo-order: chain, diamond, cycle→None, empty→Some(vec[])). Clippy clean onlance-graph-contract.🤖 Generated with Claude Code
Generated by Claude Code