diff --git a/.claude/board/AGENT_LOG.md b/.claude/board/AGENT_LOG.md index 5e027ce27..40bf56722 100644 --- a/.claude/board/AGENT_LOG.md +++ b/.claude/board/AGENT_LOG.md @@ -1,3 +1,9 @@ +## 2026-10-16 — N21 verification (guarded-executor, Haiku) + +- One card: #1570's five commands with `CARGO_PROFILE_*_DEBUG=0`, `CARGO_INCREMENTAL=0`. All exit 0; no retry used. +- Spot check: the `haspelmath` test count (6 passed, 196 filtered) re-run on the main thread, identical. +- Misses: none. Written by the orchestrating main thread only. + ## 2026-10-16 — synergy map since 2026-10-11 (five Haiku harvest cards) - substrate-mapper ×3 (merge maps for lance-graph and ndarray/OGAR/ruff; board delta), quote-harvester ×1 (13 stale sites), census-counter ×1 (ndarray counters). Read-only; outputs in the session scratchpad. diff --git a/.claude/board/LATEST_STATE.md b/.claude/board/LATEST_STATE.md index 3758bb621..e03a2880e 100644 --- a/.claude/board/LATEST_STATE.md +++ b/.claude/board/LATEST_STATE.md @@ -1,3 +1,8 @@ +## 2026-10-16 — main after #1578 (merged `1f5322fe`) and #1579 (merged `7302f58c`); ledger update in PR #1581 + +- **On main:** #1578 (ledger revision after #1576, arc catch-up), #1579 (synergy map since 2026-10-11, ten STATUS rows set to Shipped, label codebook sizes measured), #1580 (hydrate: `zip` 2 → 9). +- **In PR #1581:** ledger update deduplicated against #1579; DOC-WIRE (2) stale 59..60 prose corrected; N21 verified green. Docs only. + ## 2026-10-16 — main after #1576 (merged `9a657bd3`); catch-up for #1560, #1570, #1571, #1574, #1575, #1576 - **On main, not recorded here before:** #1560 (O22 `CounterfactualVerdict` producer, R1 test, ledger correction), #1570 (Haspelmath harness, frozen operand dictionary), #1571 (D-QPROF-0 `qualia_profile`, merged; the "in PR" section below is now on main), #1574 (59..60 one field, two vocabularies; mapping inventory; D-ASM-0), #1575 (D-HOLO H14d measured), #1576 (D-ASM-1..4 spec ratified, nothing authorized). diff --git a/.claude/board/PR_ARC_INVENTORY.md b/.claude/board/PR_ARC_INVENTORY.md index d4fb40598..e9267da8e 100644 --- a/.claude/board/PR_ARC_INVENTORY.md +++ b/.claude/board/PR_ARC_INVENTORY.md @@ -1,3 +1,10 @@ +## 2026-10-16 — lance-graph PR #1578 (merged `1f5322fe`, branch `ccr-b2e415d9-4jfvyk`, head `8b669321`) — recovery ledger revised against #1560–#1576 + +- **Added:** `.claude/research/2026-10-recovery/LEDGER-REVISION-2026-10-16.md` (closed C22–C30; open items merged and deduplicated; K2', K6, K7, K8; fruits; operator decisions with recommendations); entry `2026-10-16-ledger-revision-after-1576.md`; arc entries for #1560, #1570, #1571, #1574, #1575, #1576; LATEST_STATE section. +- **Locked:** none (docs only). +- **Deferred / open:** O30, O29, DOC-WIRE (1), BAND-RETIRE, D-ASM authorization; superseded in part by `LEDGER-UPDATE-2026-10-16b.md` (#1581). +- **Confidence:** medium; three harvest-scout zero-hit claims were false and re-checked by hand. + ## 2026-10-16 — lance-graph PR #1576 (merged `9a657bd3`, branch `ccr-c09987ed-cuk6r5`, 8 commits) — council: metta-datalog-assimilation W1–W3 spec ratified (D-ASM-1..4) - **Added:** `.claude/plans/metta-datalog-assimilation-v1-council-v3.md` (executable spec, gates G1–G9); ⊘ notes on the plan's decisions A–E, G and §6; `metta-polyfill-v1.md` §5 amended; STATUS D-ASM-1..5, D-MTP-2, D-RA-3; 3 TECH_DEBT rows; entry `2026-10-16-metta-datalog-assimilation-council-v3.md`. diff --git a/.claude/board/entries/2026-10-16-ledger-update-dedup-1579.md b/.claude/board/entries/2026-10-16-ledger-update-dedup-1579.md new file mode 100644 index 000000000..f76dd1446 --- /dev/null +++ b/.claude/board/entries/2026-10-16-ledger-update-dedup-1579.md @@ -0,0 +1,17 @@ +# 2026-10-16 — Recovery ledger deduplicated against #1579; DOC-WIRE (2) and N21 closed + +**Status:** WORKING-MODEL (research document; no code changed). + +`.claude/research/2026-10-recovery/LEDGER-UPDATE-2026-10-16b.md` maps the +overlap between the 2026-10-16 ledger revision (#1578) and #1579's synergy map, +using the ledger ids #1579 §8 adopted. + +- **Closed:** + - **C30 duplicate:** #1579 kept one #1575 arc entry. + - **DOC-WIRE (2):** three docs still described bits 59..60 as two readings; each now carries a dated correction pointing at `causal-edge/src/layout.rs`. + - **N21:** #1570's five commands all exit 0: `haspelmath` 6/6, `deepnsm-v2 --all-targets` 202 + 14, clippy `-D warnings` and fmt clean, and the example prints `SPO=(doctor, gave, letter)`. +- **Regraded:** N17 (relalg absorption) is no longer low. Its fix waits on D-RA-2 (rules into the contract set IR, D-CMR-1, or delete relalg). +- **New input:** #1579 §9 measured that no two label corpora fit one 256-entry codebook, so D-ASM-1's `LabelCodebook` should be one codebook per classid and space. N16 (operand registry) is related to that, not the same item. +- **Recorded by reference from #1579:** D2 preimage op, D6 SAG on a withdrawn base, X11, X19, D8 doc drift, L6 (#1544). X18/D7 was closed by #1579. + +OPEN: DOC-WIRE (1), BAND-RETIRE, O30, O29, D-ASM authorization, D-RA-2. diff --git a/.claude/board/entries/README.md b/.claude/board/entries/README.md index d25b5b610..ce02f0d77 100644 --- a/.claude/board/entries/README.md +++ b/.claude/board/entries/README.md @@ -25,12 +25,13 @@ index row, (3) no duplicate entry id. Checks 1 and 2 are deliberately opposite directions; the stranding this convention prevents shows up in exactly one of them, never both. -352 entries, 2026-08-06 .. 2026-10-16. +353 entries, 2026-08-06 .. 2026-10-16. | date | entry id | finding | file | |---|---|---|---| | 2026-10-16 | `synergy-map-since-1011` | | [2026-10-16-synergy-map-since-1011.md](2026-10-16-synergy-map-since-1011.md) | | 2026-10-16 | `D-ASM-1` | | [2026-10-16-metta-datalog-assimilation-council-v3.md](2026-10-16-metta-datalog-assimilation-council-v3.md) | +| 2026-10-16 | `ledger-update-dedup-1579` | | [2026-10-16-ledger-update-dedup-1579.md](2026-10-16-ledger-update-dedup-1579.md) | | 2026-10-16 | `ledger-revision-after-1576` | | [2026-10-16-ledger-revision-after-1576.md](2026-10-16-ledger-revision-after-1576.md) | | 2026-10-16 | `D-HOLO-H14d` | | [2026-10-16-d-holo-h14d-logic-mask-read.md](2026-10-16-d-holo-h14d-logic-mask-read.md) | | 2026-10-15 | `nars-metta-ce64-mapping-inventory` | | [2026-10-15-nars-metta-ce64-mapping-inventory.md](2026-10-15-nars-metta-ce64-mapping-inventory.md) | diff --git a/.claude/knowledge/causal-plane-inventory.md b/.claude/knowledge/causal-plane-inventory.md index eae24976a..4660f8d3f 100644 --- a/.claude/knowledge/causal-plane-inventory.md +++ b/.claude/knowledge/causal-plane-inventory.md @@ -59,6 +59,13 @@ model of how to fuse two readings honestly: wire and ordinal compatibility exact legacy projection intentional, and **historical factual provenance explicitly NOT guaranteed** — it refuses to backfill meaning onto old rows. +> ⊘ **Corrected 2026-10-16 (#1574, DECISION 2026-10-15).** "Two readings" is +> wrong: bits 59..60 are ONE field under two vocabularies. `TrustTexture` +> is the MUL's name for the same cell as the `CausalTopology` variant of the +> same ordinal (`causal-edge/src/layout.rs`, `TrustTexture` doc). A +> contradiction is an epistemic pothole the MUL reacts to, never a cell of +> this field. + --- ## §2 — The six semantic families (ratified; do not add a seventh from a shape) diff --git a/.claude/plans/causaledge64-mailbox-rename-soa-v1.md b/.claude/plans/causaledge64-mailbox-rename-soa-v1.md index ba78e9b92..d80022c7c 100644 --- a/.claude/plans/causaledge64-mailbox-rename-soa-v1.md +++ b/.claude/plans/causaledge64-mailbox-rename-soa-v1.md @@ -58,6 +58,14 @@ Every cognitive-state category has **two forms**: an architectural (cold, unboun One physical field. Four consumer lenses. `lance-graph-contract::mul::TrustTexture` becomes the canonical 2-bit form; planner MUL `GateDecision` is a projection; Wisdom/Staunen are projections; `thinking-engine::qualia` Staunen depth maps to the same 2 bits. +> ⊘ **Corrected 2026-10-16 (#1574, DECISION 2026-10-15).** The table above +> is superseded. Bits 59..60 are the topology coordinate under two +> vocabularies (`Crystalline`/`Direct`, `Solid`/`IndirectKnownIntermediates`, +> `Fuzzy`/`IndirectUnknownIntermediates`, `Murky`/`Unknown`); `Dissonant` and +> "Contradiction depth" are not cells of this field. A contradiction is an +> epistemic pothole the MUL reacts to (Compass). Current table: +> `crates/causal-edge/src/layout.rs`, `TrustTexture` doc. + **Closes**: `LATEST_STATE.md` "ThinkingStyle: 4 copies (contract canonical, not yet adopted)" — by reclaiming the canonical form via 8-bit-slot rename. Also closes `THINKING_ORCHESTRATION_WIRING.md` Gap 1 (Contract Not Consumed). **Why this is CPU-shaped:** the architectural ID space is unbounded; the physical slot space is bounded by attention working-set size. Eviction = "this domain/witness/style fell out of attention this session." Rebinding = "new domain entered the conversation; claim a free slot, evict LRU if needed." Per-session different rename tables = per-session different focus-of-attention. The session IS the binding. Two sessions with different rename tables interpret the same 5-bit G value as different domains — exactly how human cognitive context-switching works neurologically (re-binding attention to different active populations). diff --git a/.claude/research/2026-10-recovery/LEDGER-UPDATE-2026-10-16b.md b/.claude/research/2026-10-recovery/LEDGER-UPDATE-2026-10-16b.md new file mode 100644 index 000000000..3ff64c92b --- /dev/null +++ b/.claude/research/2026-10-recovery/LEDGER-UPDATE-2026-10-16b.md @@ -0,0 +1,66 @@ +# Ledger update, 2026-10-16 (b): deduplicated against PR #1579 + +Updates `LEDGER-REVISION-2026-10-16.md` (merged in #1578). That file is not +edited; this one supersedes the rows named below. + +**Why.** PR #1579 (another session, branch `ccr-9f3b8075-74u47l-synergy-map`, +board only) mapped 123 fleet merges against the 2026-10-11 SIMD-fleet listing +and opened while #1578 was in review. Several of its loose ends are the same +items as rows in the 2026-10-16 revision under different names. This file +names each overlap once, keeps one id per item, and records what #1579 adds +that the revision did not have. Read from #1579 as merged (`7302f58c`). Its last commit (`97b744ec`) already +reconciled the synergy map against the 2026-10-16 revision (§8 of +`entries/2026-10-16-synergy-map-since-1011.md`) and measured label codebook +sizes (§9). Where that reconciliation names a ledger id, this file uses the +same id. + +## 1. Overlaps between the two PRs + +| here | #1579 | resolution | +|---|---|---| +| C30 (arc entry for #1575, LATEST_STATE section) | PR_ARC entry for #1575, LATEST_STATE section | Same records, written twice. #1579 resolved its conflict by keeping one #1575 arc entry (`a1b79feb`). Closed. | +| N17 relalg absorption, fruit L/H | D1 "one set algebra, four spellings", H1 (#1579 §8: "use N17") | Id kept: **N17**. **Regraded: not low.** #1579 D1 shows the fix has no agreed home until D-RA-2 decides whether relalg's rules move into the contract set IR (D-CMR-1, still Queued) or relalg is deleted. The two reproducers can still become failing tests now; the fix waits on D-RA-2. | +| N18 version clock not wired into a reader, soak outstanding | arc "WAL": reader wiring, 60-min soak truncated at 27.5 min | Same item. Keep N18; the soak figure (27.5 of 60 min) comes from #1579. | +| N19 layout costs, no A/B/C verdict | D3 read masks, clustering, prefilter; H3 S3 arm; H4 clustered write layout | Same stream. #1579 adds the operational form: bytes fall only when survivors are page-local, so the write layout must cluster by the apertures that read it (H4), and on object storage clustering cuts requests, not only bytes (H3). Keep N19 as the umbrella; H3/H4 are its next steps. | +| N20 qualia CLUT | "Research docs" arc, `ISS-QUALIA-A2B-CLUT-UNMEASURED` | Same issue row. No change. | +| N15 → D-ASM | V2 (D-ASM W1–W3, not authorized) and V1 (label codebooks) | Same item. **New input from #1579 §9 (measured):** no two corpora fit one 256-entry codebook (MeTTa stdlib alone has 266 symbols; per script at most 45), so V1 now reads "one codebook per classid and space, not one shared". D-ASM-1's `LabelCodebook` should be that per-space codebook; decide before W1 builds. | +| K2' one evidence root, O30 width | D4 truth: D-NALP-4 (per-row truth lane vs scalar law per survivor), D-HL-3 evidential base | Adjacent, not identical. O30 decides how wide a root list is; D-NALP-4 decides whether a population carries truth per row. Both must agree before revision runs over populations (D-NALP-5). Linked, kept separate. | +| N16 operand registry (`u16 WordId → u8` operand per sealed epoch) | D5 / V1 label codebooks | **Related, not the same** (#1579 §8). N16 maps word ids to 8:8:8 wire operands per epoch; V1 maps symbols to label ids per classid and space. Shared shape: both are frozen, version-identified codebooks selected by scope, and the 256 bound in #1579 §9 is the same `u8` operand width N16 targets. Kept as two items. | + +## 2. Rows from #1579 not in the 2026-10-16 revision + +Recorded here by reference so this ledger stays complete; #1579 owns the +detail. + +| id | item | where | +|---|---|---| +| D2 | One preimage operation (run set under a monotone map) serves D-CMR-5, D-ASM-2, D-RA-4 and Gremlin `out()`; the D-MTP-1b probe is its oracle | #1579 §4 D2, fruit H2 | +| D6 | D-SAG-5-1..10 are queued against D-SAG-1, which #1524 withdrew; they wait on the D-CMR rebuild | #1579 §4 D6, V4 | +| X11 | `F32x8::mul_add` is documented as fused on every arm; the wasm arm without `relaxed-simd` does two roundings | #1579 §1, L4 (ndarray) | +| X19 | lgj config comment stale; `check_build_cpu` not called in the handshake | #1579 L5 (lance-graph-java) | +| D8 | Doc drift that contradicts build rulings (`.cargo/config.toml:55-56`, tesseract and lichess toolchain prose, TD-POPCNT item (b)) | #1579 L3 | +| L6 | #1544 open only as a reminder of D-NALP-4; close with an ISSUES row | #1579 L6 | +| STATUS sweep | ten rows still read "In PR" after their merges | done in #1579 | +| X18 / D7 | Two `ndarray` package ids | closed in #1579: the root and relalg `[patch]` sections resolve OGAR's git ndarray to the same `../ndarray` | + +## 3. Done in this PR + +| id | what | evidence | +|---|---|---| +| DOC-WIRE (2) | The three stale lines that still described bits 59..60 as two readings now carry a dated correction pointing at `causal-edge/src/layout.rs`'s `TrustTexture` doc: `.claude/knowledge/causal-plane-inventory.md` (after the two-readings paragraph), `docs/TYPE_DUPLICATION_MAP.md` (row 1), `.claude/plans/causaledge64-mailbox-rename-soa-v1.md` (after the collapse table). The originals are kept, marked superseded. | this PR | +| N21 | The five verification commands from #1570's body, run. All exit 0: `haspelmath` 6/6; `deepnsm-v2 --all-targets` 202 + 14 passed; clippy `-D warnings` and fmt clean; the example prints `SPO=(doctor, gave, letter) ce64=0000000000020100 map=Connected`. | guarded-executor run; the haspelmath count re-run on the main thread | + +DOC-WIRE (1) (one global reading of 43..52 in +`docs/architecture/ce64-semantic-upper-half.md`) is not touched: it waits on +the operator's yes. + +## 4. Open items after this update + +Unchanged from the 2026-10-16 revision unless named above: O25 + K2' (needs +O30), O30, K6 (D-HL-5 H4), O29, D-46, D-MATURATION3-1, ABL, D-ASM (now +with the per-space codebook input), O23, R2–R4, O26, O7/O13/O15, DOC-WIRE +(1), BAND-RETIRE, N16 (related to V1), N18, N19, N20, and N17 (gated on +D-RA-2). Closed by this update: C30 duplicate, DOC-WIRE (2), N21. + +Low-hanging left in this ledger's own scope: R3/R4 wording, and the two +absorption reproducers as failing tests (the fix waits on D-RA-2). diff --git a/docs/TYPE_DUPLICATION_MAP.md b/docs/TYPE_DUPLICATION_MAP.md index e714453f6..b28b6fde7 100644 --- a/docs/TYPE_DUPLICATION_MAP.md +++ b/docs/TYPE_DUPLICATION_MAP.md @@ -12,7 +12,7 @@ Four enums with the same name, none importing any other, encoding **different co | # | Location | Line | Variants | Arity | Semantic axis | |---|----------|------|----------|-------|---------------| -| 1 | `crates/causal-edge/src/layout.rs` | 141 | Crystalline / Solid / Fuzzy / Murky | 4 | Pearl-3 epistemic lens; discriminator = 2-bit truth field in CausalEdge64 (bits 59-60) | +| 1 | `crates/causal-edge/src/layout.rs` | 141 | Crystalline / Solid / Fuzzy / Murky | 4 | MUL vocabulary for the topology coordinate, bits 59-60 of CausalEdge64: the same cell as `CausalTopology` of the same ordinal, not a second lens (⊘ was "Pearl-3 epistemic lens"; corrected 2026-10-16 per #1574) | | 2 | `crates/lance-graph-contract/src/mul.rs` | 82 | Calibrated / Overconfident / Uncertain / Underconfident | 4 | MUL meta-uncertainty layer (felt vs demonstrated competence) | | 3 | `crates/lance-graph-planner/src/mul/trust.rs` | 30 | Crystalline / Solid / Fuzzy / Murky / Dissonant | 5 | MUL trust-quality axis; **5 variants are unrepresentable in the 2-bit CausalEdge64 truth field** — this one can never be the carrier-side type | | 4 | `crates/lance-graph/src/graph/arigraph/orchestrator.rs` | 114 | Crystalline / Fibrous / Fuzzy | 3 | AriGraph topology-weight reliability lens |