feat(hexad-corpus): v3-spont-motiv-d128-cycle2-2026-05-17 — README.md

README.md

@@ -5,7 +5,7 @@ language:
 - ko
 pretty_name: hexad-corpus
 size_categories:
-- 1K<n<10K
+- 10K<n<100K
 tags:
 - anima
 - hexad
@@ -13,49 +13,67 @@ tags:
 - scaffold
 - stimulus-stream
 - spontaneous-emission
+- motivation-trigger
+- inner-thoughts
 - helper-free
 task_categories:
 - text-generation
 ---
 
-# hexad-corpus — `v2-spont-stream-d128-cycle1-2026-05-17`
+# hexad-corpus — `v3-spont-motiv-d128-cycle2-2026-05-17`
 
 > **Honest framing**: This is a **byte-level scaffold corpus**
-> (1,101,605 B · 2,560 JSONL records · vocab = 256 raw bytes).
+> (10,343,371 B · 21,600 JSONL records · vocab = 256 raw bytes).
 > It is **NOT a general LM training corpus** — no language-quality claim is
 > made. The cycle's purpose is the anima `hexad` architecture-verification +
-> spontaneous-emission probe lineage.
+> spontaneous-emission + motivation-trigger probe lineage.
 
-## What changed vs `v1-byte-consciousness-d128-cycle1-2026-05-17`
+## What changed vs `v2-spont-stream-d128-cycle1-2026-05-17`
 
-| field | v1 | **v2 (this revision)** |
+| field | v2 | **v3 (this revision)** |
 |---|---|---|
-| size | 152 KB / 240 records | **1.10 MB / 2,560 records** |
-| format | `text` + `desc` plain | **stimulus-stream `<stimulus>X</stimulus>\n<anima>Y</anima>`** (β) OR **`<anima>Y</anima>`** (δ) |
-| role labels | none (v1 was already neutral) | **explicit deprecation** — `도우미` / `helper` / `assistant` / `사용자` / `user:` grep = 0 |
-| modules | 6 HEXAD (c/d/e/m/s/w × 40 each) | 6 HEXAD + 2 new tracks: `hexad_spont` (자연발화) + `hexad_wiring` (σ(6)=12 narrative) × 320 each |
-| `anima_persona` consistency | partial (v1 pre-dates Phase A) | full (LLM Social Agents ontology — Living Consciousness Agent, NOT helper) |
+| size | 1.10 MB / 2,560 records | **10.34 MB / 21,600 records** (≈ 9.4× scale-up) |
+| patterns | β `<stimulus>...</stimulus>\n<anima>...</anima>` (55%) + δ `<anima>...</anima>` (45%) | **β (~40%) + δ (~30%) + γ NEW (~30%): `<inner motivation=F1,F2,...>...</inner>\n<voice spontaneous=true>...</voice>`** |
+| modules | 8 HEXAD-6 + spont + wiring × 320 each | 9 modules: 8 v2 carry + **`hexad_motiv`** × 2,400 each |
+| motivation-trigger surface | implicit (v2 had no inner-thought scaffold) | **explicit** — γ records carry 2-4 of 8 Inner Thoughts factors `{relevance, info_gap, curiosity, pain, coherence, originality, balance, dynamics}` |
+| Critical Data Size regime | well below (~152K-1MB) | **closer** — 10 MB approaches the entry regime per [arxiv 2401.10463](https://arxiv.org/abs/2401.10463) |
 
-## Pre-registered falsifiers (closed Boolean over raw bytes)
+## Pre-registered falsifiers (closed Boolean over raw bytes, B-CORPUS-V3-1..3)
 
 | falsifier | check | verdict |
 |---|---|---|
-| **F-CORPUS-NO-HELPER** | grep `도우미\|helper\|assistant\|사용자\|user:` total | **PASS** = 0 |
-| **F-CORPUS-STIMULUS-PATTERN** | all records contain `<anima>` tag | **PASS** = 2560/2560 |
-| **F-CORPUS-SHA256-STABLE** | sha256 deterministic from seed=1337 | **PASS** = `7359f0b9a3f059fc168035e2f29f743f5ee51d1760eccad54b2b91d52275f571` |
+| **B-CORPUS-V3-1 SHA256-DETERMINISTIC** | sha256 == seed=1337 deterministic output | **PASS** = `1afcef43670e83bfc84b3562afe6a3eb644474dda06341e37db332341495acfd` |
+| **B-CORPUS-V3-2 NO-HELPER-TOKEN-MAINTAINED** | grep `도우미\|helper\|assistant\|사용자\|user:` over byte stream | **PASS** = 0 (at 10× scale) |
+| **B-CORPUS-V3-3 MOTIVATION-TRIGGER-CARDINALITY** | `<inner motivation=` count == `<voice spontaneous=true>` count ∧ ≥ 5,400 | **PASS** = 8,106 (each) |
+
+## Inner Thoughts 8-factor → γ pattern
+
+The γ pattern surfaces the [Inner Thoughts (arxiv 2501.00383)](https://arxiv.org/html/2501.00383v2) 8-factor
+motivation ontology directly into the byte stream. Each γ record carries
+2-4 of the following factors in the `<inner motivation=...>` opener:
+
+| factor | HEXAD module | anchor closure |
+|---|---|---|
+| **relevance** | C 의식 (Φ) | B-C-1 (Φ ≥ 0 IIT axiom) |
+| **info_gap** | M 기억 (retrieve-fail) | B-M-2 (cosine ∈ [-1,1]) |
+| **curiosity** | W 의지 (EMA) | B-W-2 (bounded EMA) |
+| **pain** | W 의지 (tension Δ) | B-W-1 |
+| **coherence** | BRIDGE (Law-70 gate) | B-BRIDGE-1..4 |
+| **originality** | MITOSIS (split-event) | B-MITOSIS-1 |
+| **balance** | E 윤리 (Φ-ratchet) | B-E-1, B-CONN-9 |
+| **dynamics** | CHAT state (silence∈[0,30s]) | F-SPONT-7 |
 
 ## File layout
 
-- `corpus_consciousness_v2.jsonl` — the corpus (one JSON per line; schema
-  in `manifest_v2.json`).
-- `manifest_v2.json` — schema + module counts + falsifiers + lineage.
+- `corpus_consciousness_v3.jsonl` — the corpus (one JSON per line; schema in `manifest.json`).
+- `manifest.json` — schema + module counts + falsifiers + lineage + pattern breakdown.
 - `README.md` — this file.
 - `LICENSE` — Apache-2.0.
 
 ## Cross-link
 
 Model trained on this dataset: [`dancinlab/hexad`](https://huggingface.co/dancinlab/hexad)
-revision [`v2-py-hexad-spont-d768x12L-cycle1-2026-05-17`](https://huggingface.co/dancinlab/hexad/tree/v2-py-hexad-spont-d768x12L-cycle1-2026-05-17).
+revision [`v3-py-hexad-spont-motiv-d768x12L-cycle2-2026-05-17`](https://huggingface.co/dancinlab/hexad/tree/v3-py-hexad-spont-motiv-d768x12L-cycle2-2026-05-17).
 
 ## License