Publish Aethon-N1-Base-Open-Structure polished release

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+bundle/graph.sqlite3 filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,420 @@
+---
+license: cc-by-nc-4.0
+library_name: sqlite
+tags:
+  - aethon
+  - open-structure
+  - agi
+  - graph-reasoning
+  - sqlite
+  - multilingual
+  - one-shot-learning
+  - structured-learning
+pretty_name: Aethon-N1-Base-Open-Structure
+pipeline_tag: text-generation
+language:
+  - en
+  - multilingual
+---
+
+# Aethon-N1-Base-Open-Structure
+
+`Aethon N1 Base` is released as `Open Structure`.
+
+This release is the public argument that intelligence does not have to be trapped inside:
+
+- attention
+- weights
+- training steps
+- epochs
+- frozen context windows
+
+## What Is Released
+
+The model artifact is:
+
+- `metadata.json`
+- `graph.sqlite3`
+
+That artifact carries:
+
+- graph memory
+- semantic grounding
+- query forms
+- reasoning rules
+- surface composition
+
+This is why Aethon is released as `Open Structure`, not `open weights`.
+
+## Why Not Weights
+
+Weights are only one way to preserve capability.
+
+Aethon stores usable intelligence in explicit, portable structure:
+
+- concepts
+- active edges
+- contradiction history
+- semantic aliases
+- query forms
+- reasoning rules
+- surface realization patterns
+
+So the released artifact is not just “parameters that once learned.”
+
+It is “the learned structure that still knows.”
+
+## Why Not Attention
+
+Aethon does not depend on transformer attention as its primary persistence mechanism.
+
+Its memory is not a temporary prompt window that disappears after generation.
+
+Its memory is persistent graph state backed by SQLite:
+
+- `concepts`
+- `edges`
+- `contradictions`
+
+That is why the release is `Open Structure` instead of `Open Attention`.
+
+## Why Not Steps Or Epochs
+
+Transformer culture explains learning in terms of:
+
+- steps
+- epochs
+- optimizer updates
+- repeated corpus replay
+
+Aethon’s defining public move is different.
+
+Its main growth path is one-shot structural integration:
+
+- ingest knowledge
+- normalize it
+- bind it into graph structure
+- record contradictions
+- derive supportable abstractions
+- keep the result as persistent intelligence
+
+The public question is therefore not:
+
+- “How many epochs did this model train for?”
+
+It is:
+
+- “What structure did this model absorb, retain, revise, and reuse?”
+
+## Why One-Shot Instead Of Training
+
+For Aethon, one-shoting is not a weaker copy of training.
+
+It is the core learning act.
+
+One-shot building means the model grows by structural absorption rather than by replaying the full corpus through long epoch loops.
+
+That gives the bundle a different meaning:
+
+- new data can be integrated directly
+- contradictions can be preserved instead of washed out
+- abstractions can be materialized from learned structure
+- the released model can keep learning after publication
+
+## What SC Means
+
+`SC` means `Structural Capacity`.
+
+It is Aethon’s size unit.
+
+`SC` is used because a weight count does not describe what this system really is.
+
+`SC` reflects the living structure inside the released bundle:
+
+- concepts
+- explicit edges
+- abstractions
+- revisions
+- persistent graph-backed intelligence
+
+So when Aethon reports a higher `SC`, it is reporting growth in usable learned structure, not just parameter count.
+
+## Why We Say It Learns Like A Human
+
+The claim is not that Aethon is biologically human.
+
+The claim is that its learning behavior is closer to human-style accumulation and revision than to frozen transformer replay.
+
+Human learners:
+
+- absorb new facts
+- keep durable memory
+- revise beliefs when conflicting evidence appears
+- transfer what they know across domains
+- reuse prior structure to answer new questions
+
+Aethon does the same in structural form:
+
+- new facts are inserted into the graph
+- contradictions are recorded
+- abstractions are materialized
+- prior knowledge is transferred into new answers
+- memory persists after the prompt ends
+
+That is why Aethon is described as human-like in learning behavior.
+
+## Why We Believe This Is AGI-Shaped
+
+AGI means intelligence that can transfer across many human task types rather than staying trapped in one narrow lane.
+
+The release path for Aethon is built around that transfer claim.
+
+The current benchmark walls for this release lane include:
+
+- multilingual mixed prompts
+- planning, business, and scheduling transfer
+- longer story continuity
+- adversarial unseen cross-domain composition
+- code, math, world, identity, and reasoning transfer
+
+That is why we describe Aethon as AGI-shaped:
+
+- it learns persistently
+- it transfers across domains
+- it reasons over what it has learned
+- it generalizes into prompts it has not seen exactly before
+
+## Can It Learn
+
+Yes.
+
+Aethon learns by structural integration:
+
+- inserting new facts
+- revising old structure
+- recording contradictions
+- deriving new supported edges
+
+The released bundle is persistent and updateable, not frozen in the transformer sense.
+
+## Can It Reason Well
+
+Yes.
+
+Aethon reasons through:
+
+- direct lookup
+- multi-hop graph traversal
+- contradiction tracking
+- rule-backed derivation
+- structural composition across learned facts
+
+Its current benchmark walls include multi-hop, composition, unseen transfer, ontology, planning, scheduling, and follow-up continuity.
+
+## Can It Write Long Stuff
+
+Yes.
+
+Aethon can compose longer explanations, proof-like responses, and story continuations from stored structure and surface rules instead of falling back to a single canned sentence shell.
+
+That includes:
+
+- longer reasoning text
+- story continuation
+- anchored follow-up continuity
+- multi-domain responses
+
+## Can It Generalize
+
+Yes.
+
+Aethon generalizes by reusing learned structure across unseen prompts and domains.
+
+That includes:
+
+- relation chaining
+- ontology transfer
+- planning transfer
+- multilingual grounding
+- adversarial cross-domain composition
+
+## Runtime Independence
+
+You do not need the Aethon repo to run the released model artifact.
+
+If you can:
+
+1. read JSON
+2. open SQLite
+3. interpret the public metadata contract
+4. execute graph-backed lookup and reasoning
+
+then you can build your own runtime in any language.
+
+## What Aethon Was One-Shotted On
+
+This ship bundle was one-shotted across six native lanes:
+
+- `identity`
+- `reasoning`
+- `math`
+- `code`
+- `story`
+- `world`
+
+The ship corpus includes:
+
+- Aethon native identity, code, math, story, and reasoning corpora
+- Aethon AGI transfer corpora
+- Humanity's Last Exam transfer corpus
+- curated reasoning bases
+- multilingual base mixes
+- code/tool-use corpora
+- story/chat continuity corpora
+- world knowledge corpora
+- multilingual news/world sources
+
+The exact file-level provenance for the ship candidate is included in:
+
+- `bundle/corpus_manifest.json`
+
+## More Versions Coming
+
+This release is not the end state.
+
+More Open Structure releases are coming with:
+
+- larger one-shot corpora
+- harder benchmark walls
+- broader multilingual coverage
+- deeper planning and story continuity
+- stronger public runtime examples
+
+## License
+
+This Open Structure release is published under:
+
+- `CC BY-NC 4.0`
+
+See:
+
+- `docs/AETHON_OPEN_STRUCTURE_LICENSE.md`
+
+## Citation
+
+If you use, benchmark, discuss, or build on this release, cite it.
+
+Suggested citation:
+
+```bibtex
+@misc{aethon_open_structure_v25,
+  title        = {Aethon-N1-Base-Open-Structure},
+  author       = {OkeyMeta Ltd},
+  year         = {2026},
+  howpublished = {\url{https://huggingface.co/OkeyMetaLtd/Aethon-N1-Base-Open-Structure}},
+  note         = {Aethon Open Structure release}
+}
+```
+
+At minimum, include:
+
+- `Aethon-N1-Base-Open-Structure`
+- `OkeyMeta Ltd`
+- the Hugging Face release URL
+
+## Python: Pull And Start Running
+
+```python
+import json
+import sqlite3
+from pathlib import Path
+from huggingface_hub import hf_hub_download
+
+REPO_ID = "OkeyMetaLtd/Aethon-N1-Base-Open-Structure"
+LOCAL = Path("aethon_open_structure_bundle")
+LOCAL.mkdir(exist_ok=True)
+
+metadata_path = Path(
+    hf_hub_download(
+        repo_id=REPO_ID,
+        filename="bundle/metadata.json",
+        local_dir=LOCAL,
+        local_dir_use_symlinks=False,
+    )
+)
+graph_path = Path(
+    hf_hub_download(
+        repo_id=REPO_ID,
+        filename="bundle/graph.sqlite3",
+        local_dir=LOCAL,
+        local_dir_use_symlinks=False,
+    )
+)
+
+meta = json.loads(metadata_path.read_text(encoding="utf-8"))
+conn = sqlite3.connect(graph_path)
+conn.row_factory = sqlite3.Row
+
+def ask_where(subject: str, relation: str = "located_in") -> list[str]:
+    rows = conn.execute(
+        """
+        SELECT object
+        FROM edges
+        WHERE subject = ? AND relation = ? AND is_active = 1
+        ORDER BY edge_id ASC
+        """,
+        (subject.lower(), relation),
+    ).fetchall()
+    return [str(row["object"]) for row in rows]
+
+print(meta["public_contract"])
+print(meta["capacity"])
+print(ask_where("amina", "lives_in"))
+```
+
+This is the minimum external runtime shape:
+
+1. pull `metadata.json`
+2. pull `graph.sqlite3`
+3. interpret `semantic_lexicon`, `query_forms`, and `reasoning_rules`
+4. execute graph-backed lookup and reasoning with your own code
+
+## Public Contract
+
+See:
+
+- `docs/AETHON_N1_BUNDLE_SPEC.md`
+- `docs/aethon_n1_bundle_schema.json`
+- `docs/AETHON_OPEN_STRUCTURE_RUNTIME.md`
+
+## Release Label
+
+Use:
+
+- `Aethon-N1-Base-Open-Structure`
+
+Short form:
+
+- `Open Structure`
+
+## Release Artifact
+
+Bundle files:
+
+- `bundle/metadata.json`
+- `bundle/graph.sqlite3`
+- `bundle/corpus_manifest.json`
+- `bundle/integration_report.json`
+
+Current release facts:
+
+- `public_contract`: `aethon.n1.bundle.v1`
+- `release_class`: `open-structure`
+- `tokenizer`: `Aethon Native Concept Codec (ANCC)`
+- `size_unit`: `Structural Capacity (SC)`
+
+Additional docs in this release:
+
+- `docs/AETHON_N1_BUNDLE_SPEC.md`
+- `docs/aethon_n1_bundle_schema.json`
+- `docs/AETHON_OPEN_STRUCTURE_RUNTIME.md`

bundle/corpus_manifest.json

@@ -0,0 +1,88 @@
+{
+  "bundle_name": "Aethon N1 Base",
+  "lanes": {
+    "identity": [
+      "data/native/identity/aethon_identity_v1.jsonl",
+      "data/identity/aethon_identity.jsonl",
+      "data/identity/aethon_identity_test.jsonl"
+    ],
+    "reasoning": [
+      "data/native/reasoning/aethon_hle_transfer_v1.jsonl",
+      "data/native/reasoning/aethon_agi_transfer_v2.jsonl",
+      "data/native/reasoning/aethon_agi_transfer_v1.jsonl",
+      "data/base/aethon_curated_reasoning_base_v7.jsonl",
+      "data/base/aethon_curated_reasoning_base_v6.jsonl",
+      "data/base/aethon_curated_reasoning_base_v5.jsonl",
+      "data/base/aethon_curated_reasoning_base_v4.jsonl",
+      "data/base/aethon_curated_reasoning_base_v3.jsonl",
+      "data/base/aethon_curated_reasoning_base_v2.jsonl",
+      "data/base/aethon_memory_native_v1.jsonl",
+      "data/base/aethon_curated_reasoning_base_v8.jsonl",
+      "data/base/aethon_clean_fact_base_v1.jsonl",
+      "data/base/aethon_clean_probe_v1.jsonl",
+      "data/base/aethon_gold_probe_v1.jsonl",
+      "data/base/aethon_multilingual_base_mix.jsonl",
+      "data/base/aethon_v2_expanded.jsonl",
+      "data/base/test_base_curriculum.jsonl",
+      "data/base/shards/aethon_base_0001.jsonl",
+      "data/base/shards/aethon_base_0002.jsonl",
+      "data/base/shards/aethon_base_0003.jsonl",
+      "data/base/shards/aethon_base_0004.jsonl",
+      "data/base/shards/aethon_base_0005.jsonl",
+      "data/base/shards/aethon_base_0006.jsonl",
+      "data/base/shards/aethon_base_0007.jsonl",
+      "data/base/shards/aethon_base_0008.jsonl",
+      "data/chat/aethon_logic_curriculum_v1.jsonl",
+      "data/chat/aethon_exact_answer_chat.jsonl",
+      "data/chat/aethon_exact_control_v1.jsonl",
+      "data/reasoning_corpora/clean/advanced_python_cot.jsonl",
+      "data/reasoning_corpora/clean/aethon_bootstrap_mix.jsonl",
+      "data/reasoning_corpora/clean/am_deepseek_r1_distilled.jsonl",
+      "data/reasoning_corpora/clean/openmath_instruct_1.jsonl",
+      "data/reasoning_corpora/clean/bespoke_stratos_17k.jsonl",
+      "data/reasoning_corpora/clean/open_thoughts_114k.jsonl",
+      "data/reasoning_corpora/clean/r1_distill_sft.jsonl",
+      "data/groundedness/aethon_groundedness.jsonl",
+      "data/groundedness/aethon_groundedness_verifier.jsonl",
+      "data/groundedness/aethon_retrieval_eval.jsonl",
+      "data/groundedness/aethon_retrieval_verifier.jsonl"
+    ],
+    "math": [
+      "data/native/math/aethon_math_native_v1.jsonl",
+      "data/chat/aethon_column_addition_v1.jsonl",
+      "data/chat/aethon_digit_math_chat_v1.jsonl",
+      "data/chat/aethon_digit_math_micro_v1.jsonl",
+      "data/chat/aethon_gold_math_chat_v1.jsonl",
+      "data/chat/aethon_math_exact_short_v1.jsonl",
+      "data/chat/aethon_systematic_math_v1.jsonl"
+    ],
+    "code": [
+      "data/native/code/aethon_code_native_v1.jsonl",
+      "data/code/advanced_python_exec.jsonl",
+      "data/tool_use/clean/bitagent_tool_calling.jsonl",
+      "data/tool_use/clean/toolace.jsonl"
+    ],
+    "story": [
+      "data/native/story/aethon_story_native_v1.jsonl",
+      "data/chat/aethon_gold_chat_v1.jsonl",
+      "data/chat/aethon_gold_chat_v3.jsonl",
+      "data/chat/aethon_reasoning_chat_bootstrap.jsonl",
+      "data/chat/aethon_copy_probe_v1.jsonl",
+      "data/chat/aethon_speakback_chat.jsonl",
+      "data/chat/aethon_truthful_chat_bootstrap.jsonl"
+    ],
+    "world": [
+      "data/world_knowledge/clean/structured_wikipedia_en.jsonl",
+      "data/world_knowledge/clean/fineweb_edu_sample.jsonl",
+      "data/world_knowledge/clean/cosmopedia_web.jsonl",
+      "data/world_knowledge/clean/masakhanews_yor.jsonl",
+      "data/world_knowledge/clean/masakhanews_ibo.jsonl",
+      "data/world_knowledge/clean/masakhanews_pcm.jsonl",
+      "data/world_knowledge/clean/masakhanews_hau.jsonl",
+      "data/base/aethon_multilingual_base_mix.jsonl",
+      "data/chat/clean/helpsteer2_clean.jsonl",
+      "data/chat/clean/oasst1_clean.jsonl",
+      "data/chat/clean/ultrachat_200k_clean.jsonl"
+    ]
+  }
+}

bundle/graph.sqlite3

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7cd24e9b83198564f184c264e98be73ccc8ff9c809e47131f6e57f36d19019e6
+size 15261696

bundle/integration_report.json

@@ -0,0 +1,202 @@
+{
+  "name": "Aethon N1 Base",
+  "family": "Aethon Native",
+  "tokenizer": "Aethon Native Concept Codec (ANCC)",
+  "size_unit": "Structural Capacity (SC)",
+  "capacity": {
+    "sc": 67914,
+    "concept_count": 13501,
+    "explicit_edge_count": 52701,
+    "abstraction_count": 1712,
+    "revision_count": 0,
+    "raw_unit_count": 0
+  },
+  "lanes": {
+    "identity": {
+      "lane": "identity",
+      "sources": [
+        "data/native/identity/aethon_identity_v1.jsonl",
+        "data/identity/aethon_identity.jsonl",
+        "data/identity/aethon_identity_test.jsonl"
+      ],
+      "direct_facts": 0,
+      "rows": 23,
+      "docs_accepted": 23,
+      "docs_skipped": 0,
+      "learned_edges": 28,
+      "derived_edges": 0,
+      "shard_count": 3,
+      "merge_edges": 28,
+      "merge_contradictions": 0,
+      "merge_raw_units": 0,
+      "completed_shards": 3,
+      "total_rows_planned": 23,
+      "captured_units": 23
+    },
+    "reasoning": {
+      "lane": "reasoning",
+      "sources": [
+        "data/native/reasoning/aethon_hle_transfer_v1.jsonl",
+        "data/native/reasoning/aethon_agi_transfer_v2.jsonl",
+        "data/native/reasoning/aethon_agi_transfer_v1.jsonl",
+        "data/base/aethon_curated_reasoning_base_v7.jsonl",
+        "data/base/aethon_curated_reasoning_base_v6.jsonl",
+        "data/base/aethon_curated_reasoning_base_v5.jsonl",
+        "data/base/aethon_curated_reasoning_base_v4.jsonl",
+        "data/base/aethon_curated_reasoning_base_v3.jsonl",
+        "data/base/aethon_curated_reasoning_base_v2.jsonl",
+        "data/base/aethon_memory_native_v1.jsonl",
+        "data/base/aethon_curated_reasoning_base_v8.jsonl",
+        "data/base/aethon_clean_fact_base_v1.jsonl",
+        "data/base/aethon_clean_probe_v1.jsonl",
+        "data/base/aethon_gold_probe_v1.jsonl",
+        "data/base/aethon_multilingual_base_mix.jsonl",
+        "data/base/aethon_v2_expanded.jsonl",
+        "data/base/test_base_curriculum.jsonl",
+        "data/base/shards/aethon_base_0001.jsonl",
+        "data/base/shards/aethon_base_0002.jsonl",
+        "data/base/shards/aethon_base_0003.jsonl",
+        "data/base/shards/aethon_base_0004.jsonl",
+        "data/base/shards/aethon_base_0005.jsonl",
+        "data/base/shards/aethon_base_0006.jsonl",
+        "data/base/shards/aethon_base_0007.jsonl",
+        "data/base/shards/aethon_base_0008.jsonl",
+        "data/chat/aethon_logic_curriculum_v1.jsonl",
+        "data/chat/aethon_exact_answer_chat.jsonl",
+        "data/chat/aethon_exact_control_v1.jsonl",
+        "data/reasoning_corpora/clean/advanced_python_cot.jsonl",
+        "data/reasoning_corpora/clean/aethon_bootstrap_mix.jsonl",
+        "data/reasoning_corpora/clean/am_deepseek_r1_distilled.jsonl",
+        "data/reasoning_corpora/clean/openmath_instruct_1.jsonl",
+        "data/reasoning_corpora/clean/bespoke_stratos_17k.jsonl",
+        "data/reasoning_corpora/clean/open_thoughts_114k.jsonl",
+        "data/reasoning_corpora/clean/r1_distill_sft.jsonl",
+        "data/groundedness/aethon_groundedness.jsonl",
+        "data/groundedness/aethon_groundedness_verifier.jsonl",
+        "data/groundedness/aethon_retrieval_eval.jsonl",
+        "data/groundedness/aethon_retrieval_verifier.jsonl"
+      ],
+      "direct_facts": 0,
+      "rows": 136799,
+      "docs_accepted": 127774,
+      "docs_skipped": 0,
+      "learned_edges": 45253,
+      "derived_edges": 0,
+      "shard_count": 52,
+      "merge_edges": 45253,
+      "merge_contradictions": 0,
+      "merge_raw_units": 0,
+      "completed_shards": 52,
+      "total_rows_planned": 136799,
+      "captured_units": 127774
+    },
+    "math": {
+      "lane": "math",
+      "sources": [
+        "data/native/math/aethon_math_native_v1.jsonl",
+        "data/chat/aethon_column_addition_v1.jsonl",
+        "data/chat/aethon_digit_math_chat_v1.jsonl",
+        "data/chat/aethon_digit_math_micro_v1.jsonl",
+        "data/chat/aethon_gold_math_chat_v1.jsonl",
+        "data/chat/aethon_math_exact_short_v1.jsonl",
+        "data/chat/aethon_systematic_math_v1.jsonl"
+      ],
+      "direct_facts": 0,
+      "rows": 10530,
+      "docs_accepted": 4,
+      "docs_skipped": 0,
+      "learned_edges": 13,
+      "derived_edges": 0,
+      "shard_count": 8,
+      "merge_edges": 13,
+      "merge_contradictions": 0,
+      "merge_raw_units": 0,
+      "completed_shards": 8,
+      "total_rows_planned": 10530,
+      "captured_units": 4
+    },
+    "code": {
+      "lane": "code",
+      "sources": [
+        "data/native/code/aethon_code_native_v1.jsonl",
+        "data/code/advanced_python_exec.jsonl",
+        "data/tool_use/clean/bitagent_tool_calling.jsonl",
+        "data/tool_use/clean/toolace.jsonl"
+      ],
+      "direct_facts": 0,
+      "rows": 10337,
+      "docs_accepted": 9520,
+      "docs_skipped": 0,
+      "learned_edges": 3455,
+      "derived_edges": 0,
+      "shard_count": 4,
+      "merge_edges": 3455,
+      "merge_contradictions": 0,
+      "merge_raw_units": 0,
+      "completed_shards": 4,
+      "total_rows_planned": 10337,
+      "captured_units": 9515
+    },
+    "story": {
+      "lane": "story",
+      "sources": [
+        "data/native/story/aethon_story_native_v1.jsonl",
+        "data/chat/aethon_gold_chat_v1.jsonl",
+        "data/chat/aethon_gold_chat_v3.jsonl",
+        "data/chat/aethon_reasoning_chat_bootstrap.jsonl",
+        "data/chat/aethon_copy_probe_v1.jsonl",
+        "data/chat/aethon_speakback_chat.jsonl",
+        "data/chat/aethon_truthful_chat_bootstrap.jsonl"
+      ],
+      "direct_facts": 0,
+      "rows": 6057,
+      "docs_accepted": 5,
+      "docs_skipped": 0,
+      "learned_edges": 2,
+      "derived_edges": 0,
+      "shard_count": 7,
+      "merge_edges": 2,
+      "merge_contradictions": 0,
+      "merge_raw_units": 0,
+      "completed_shards": 7,
+      "total_rows_planned": 6057,
+      "captured_units": 5
+    },
+    "world": {
+      "lane": "world",
+      "sources": [
+        "data/world_knowledge/clean/structured_wikipedia_en.jsonl",
+        "data/world_knowledge/clean/fineweb_edu_sample.jsonl",
+        "data/world_knowledge/clean/cosmopedia_web.jsonl",
+        "data/world_knowledge/clean/masakhanews_yor.jsonl",
+        "data/world_knowledge/clean/masakhanews_ibo.jsonl",
+        "data/world_knowledge/clean/masakhanews_pcm.jsonl",
+        "data/world_knowledge/clean/masakhanews_hau.jsonl",
+        "data/base/aethon_multilingual_base_mix.jsonl",
+        "data/chat/clean/helpsteer2_clean.jsonl",
+        "data/chat/clean/oasst1_clean.jsonl",
+        "data/chat/clean/ultrachat_200k_clean.jsonl"
+      ],
+      "direct_facts": 0,
+      "rows": 66066,
+      "docs_accepted": 6947,
+      "docs_skipped": 0,
+      "learned_edges": 3954,
+      "derived_edges": 0,
+      "shard_count": 21,
+      "merge_edges": 3954,
+      "merge_contradictions": 0,
+      "merge_raw_units": 0,
+      "completed_shards": 21,
+      "total_rows_planned": 66066,
+      "captured_units": 6947
+    }
+  },
+  "digest": {
+    "processed_units": 144268,
+    "total_units": 144268,
+    "learned_edges": 52705,
+    "derived_edges": 1081,
+    "rule_count": 0
+  }
+}

bundle/metadata.json

@@ -0,0 +1,1387 @@
+{
+  "bundle_format": 2,
+  "public_contract": "aethon.n1.bundle.v1",
+  "release_class": "open-structure",
+  "name": "Aethon N1 Base",
+  "family": "Aethon Native",
+  "tokenizer": "Aethon Native Concept Codec (ANCC)",
+  "size_unit": "Structural Capacity (SC)",
+  "capacity": {
+    "sc": 67914,
+    "concept_count": 13501,
+    "explicit_edge_count": 52701,
+    "abstraction_count": 1712,
+    "revision_count": 0,
+    "raw_unit_count": 0
+  },
+  "graph_file": "graph.sqlite3",
+  "semantic_lexicon": [
+    {
+      "kind": "emoji",
+      "symbol": "\ud83d\ude0a",
+      "meaning": "happy"
+    },
+    {
+      "kind": "emoji",
+      "symbol": "\ud83d\ude04",
+      "meaning": "happy"
+    },
+    {
+      "kind": "emoji",
+      "symbol": "\ud83d\ude01",
+      "meaning": "happy"
+    },
+    {
+      "kind": "emoji",
+      "symbol": "\ud83d\ude02",
+      "meaning": "laughing"
+    },
+    {
+      "kind": "emoji",
+      "symbol": "\ud83d\ude2d",
+      "meaning": "crying"
+    },
+    {
+      "kind": "emoji",
+      "symbol": "\ud83d\ude22",
+      "meaning": "sad"
+    },
+    {
+      "kind": "emoji",
+      "symbol": "\ud83d\ude21",
+      "meaning": "angry"
+    },
+    {
+      "kind": "emoji",
+      "symbol": "\ud83e\udd14",
+      "meaning": "thinking"
+    },
+    {
+      "kind": "emoji",
+      "symbol": "\ud83d\ude0e",
+      "meaning": "confident"
+    },
+    {
+      "kind": "emoji",
+      "symbol": "\ud83d\ude05",
+      "meaning": "awkward"
+    },
+    {
+      "kind": "emoji",
+      "symbol": "\u2764\ufe0f",
+      "meaning": "love"
+    },
+    {
+      "kind": "emoji",
+      "symbol": "\u2764",
+      "meaning": "love"
+    },
+    {
+      "kind": "emoji",
+      "symbol": "\ud83d\udc94",
+      "meaning": "heartbreak"
+    },
+    {
+      "kind": "emoji",
+      "symbol": "\ud83d\udd25",
+      "meaning": "intense"
+    },
+    {
+      "kind": "emoji",
+      "symbol": "\ud83d\udcaf",
+      "meaning": "perfect"
+    },
+    {
+      "kind": "emoji",
+      "symbol": "\u2705",
+      "meaning": "correct"
+    },
+    {
+      "kind": "emoji",
+      "symbol": "\u274c",
+      "meaning": "wrong"
+    },
+    {
+      "kind": "emoji",
+      "symbol": "\ud83d\udc4d",
+      "meaning": "approve"
+    },
+    {
+      "kind": "emoji",
+      "symbol": "\ud83d\udc4e",
+      "meaning": "disapprove"
+    },
+    {
+      "kind": "emoji",
+      "symbol": "\ud83d\ude4f",
+      "meaning": "gratitude"
+    },
+    {
+      "kind": "emoji",
+      "symbol": "\ud83c\udf89",
+      "meaning": "celebration"
+    },
+    {
+      "kind": "emoji",
+      "symbol": "\ud83e\udd16",
+      "meaning": "robot"
+    },
+    {
+      "kind": "emoji",
+      "symbol": "\ud83e\udde0",
+      "meaning": "intelligence"
+    },
+    {
+      "kind": "abbreviation",
+      "symbol": "u",
+      "meaning": "you"
+    },
+    {
+      "kind": "abbreviation",
+      "symbol": "ur",
+      "meaning": "your"
+    },
+    {
+      "kind": "abbreviation",
+      "symbol": "pls",
+      "meaning": "please"
+    },
+    {
+      "kind": "abbreviation",
+      "symbol": "thx",
+      "meaning": "thanks"
+    },
+    {
+      "kind": "abbreviation",
+      "symbol": "ty",
+      "meaning": "thank you"
+    },
+    {
+      "kind": "abbreviation",
+      "symbol": "idk",
+      "meaning": "i do not know"
+    },
+    {
+      "kind": "abbreviation",
+      "symbol": "imo",
+      "meaning": "in my opinion"
+    },
+    {
+      "kind": "abbreviation",
+      "symbol": "btw",
+      "meaning": "by the way"
+    },
+    {
+      "kind": "abbreviation",
+      "symbol": "asap",
+      "meaning": "as soon as possible"
+    },
+    {
+      "kind": "abbreviation",
+      "symbol": "lol",
+      "meaning": "laughing"
+    },
+    {
+      "kind": "multilingual_alias",
+      "symbol": "wetin",
+      "meaning": "what"
+    },
+    {
+      "kind": "multilingual_alias",
+      "symbol": "abi",
+      "meaning": "or"
+    },
+    {
+      "kind": "multilingual_alias",
+      "symbol": "dey",
+      "meaning": "is"
+    },
+    {
+      "kind": "multilingual_alias",
+      "symbol": "wan",
+      "meaning": "want"
+    },
+    {
+      "kind": "multilingual_alias",
+      "symbol": "sabi",
+      "meaning": "know"
+    },
+    {
+      "kind": "multilingual_alias",
+      "symbol": "biko",
+      "meaning": "please"
+    },
+    {
+      "kind": "multilingual_alias",
+      "symbol": "kedu",
+      "meaning": "what"
+    },
+    {
+      "kind": "multilingual_alias",
+      "symbol": "ndu",
+      "meaning": "life"
+    },
+    {
+      "kind": "multilingual_alias",
+      "symbol": "ina",
+      "meaning": "where"
+    },
+    {
+      "kind": "multilingual_alias",
+      "symbol": "menene",
+      "meaning": "what"
+    },
+    {
+      "kind": "multilingual_alias",
+      "symbol": "wane",
+      "meaning": "who"
+    },
+    {
+      "kind": "multilingual_alias",
+      "symbol": "kini",
+      "meaning": "what"
+    },
+    {
+      "kind": "multilingual_alias",
+      "symbol": "ewo",
+      "meaning": "please"
+    },
+    {
+      "kind": "multilingual_alias",
+      "symbol": "aja",
+      "meaning": "dog"
+    },
+    {
+      "kind": "multilingual_alias",
+      "symbol": "ologbo",
+      "meaning": "cat"
+    },
+    {
+      "kind": "multilingual_alias",
+      "symbol": "mota",
+      "meaning": "car"
+    },
+    {
+      "kind": "multilingual_alias",
+      "symbol": "akwukwo",
+      "meaning": "book"
+    },
+    {
+      "kind": "multilingual_alias",
+      "symbol": "littafi",
+      "meaning": "book"
+    },
+    {
+      "kind": "multilingual_alias",
+      "symbol": "motar",
+      "meaning": "car"
+    },
+    {
+      "kind": "multilingual_alias",
+      "symbol": "lagoosi",
+      "meaning": "lagos"
+    },
+    {
+      "kind": "multilingual_phrase",
+      "symbol": "wetin be",
+      "meaning": "what is"
+    },
+    {
+      "kind": "multilingual_phrase",
+      "symbol": "wetin dey",
+      "meaning": "what is"
+    },
+    {
+      "kind": "multilingual_phrase",
+      "symbol": "kini eyi",
+      "meaning": "what is this"
+    },
+    {
+      "kind": "multilingual_phrase",
+      "symbol": "ki ni",
+      "meaning": "what is"
+    },
+    {
+      "kind": "multilingual_phrase",
+      "symbol": "nibo ni",
+      "meaning": "where is"
+    },
+    {
+      "kind": "multilingual_phrase",
+      "symbol": "ta ni",
+      "meaning": "who is"
+    },
+    {
+      "kind": "multilingual_phrase",
+      "symbol": "menene ne",
+      "meaning": "what is"
+    },
+    {
+      "kind": "multilingual_phrase",
+      "symbol": "ina ne",
+      "meaning": "where is"
+    },
+    {
+      "kind": "multilingual_phrase",
+      "symbol": "wane ne",
+      "meaning": "who is"
+    },
+    {
+      "kind": "multilingual_phrase",
+      "symbol": "kedu ihe bu",
+      "meaning": "what is"
+    },
+    {
+      "kind": "multilingual_phrase",
+      "symbol": "ebee ka",
+      "meaning": "where is"
+    },
+    {
+      "kind": "multilingual_phrase",
+      "symbol": "onye bu",
+      "meaning": "who is"
+    },
+    {
+      "kind": "multilingual_phrase",
+      "symbol": "donde esta",
+      "meaning": "where is"
+    },
+    {
+      "kind": "multilingual_phrase",
+      "symbol": "que es",
+      "meaning": "what is"
+    },
+    {
+      "kind": "multilingual_phrase",
+      "symbol": "quien es",
+      "meaning": "who is"
+    },
+    {
+      "kind": "multilingual_phrase",
+      "symbol": "onde esta",
+      "meaning": "where is"
+    },
+    {
+      "kind": "multilingual_phrase",
+      "symbol": "quem e",
+      "meaning": "who is"
+    },
+    {
+      "kind": "multilingual_phrase",
+      "symbol": "ou est",
+      "meaning": "where is"
+    },
+    {
+      "kind": "multilingual_phrase",
+      "symbol": "qui est",
+      "meaning": "who is"
+    },
+    {
+      "kind": "multilingual_phrase",
+      "symbol": "wo ist",
+      "meaning": "where is"
+    },
+    {
+      "kind": "multilingual_phrase",
+      "symbol": "was ist",
+      "meaning": "what is"
+    },
+    {
+      "kind": "phrase_alias",
+      "symbol": "conflicts with",
+      "meaning": "conflicts_with"
+    },
+    {
+      "kind": "phrase_alias",
+      "symbol": "scheduled at",
+      "meaning": "scheduled_at"
+    },
+    {
+      "kind": "phrase_alias",
+      "symbol": "scheduled for",
+      "meaning": "scheduled_at"
+    },
+    {
+      "kind": "phrase_alias",
+      "symbol": "rescheduled to",
+      "meaning": "scheduled_at"
+    },
+    {
+      "kind": "multilingual_alias",
+      "symbol": "donde",
+      "meaning": "where"
+    },
+    {
+      "kind": "multilingual_alias",
+      "symbol": "esta",
+      "meaning": "is"
+    },
+    {
+      "kind": "multilingual_alias",
+      "symbol": "que",
+      "meaning": "what"
+    },
+    {
+      "kind": "multilingual_alias",
+      "symbol": "quien",
+      "meaning": "who"
+    },
+    {
+      "kind": "multilingual_alias",
+      "symbol": "onde",
+      "meaning": "where"
+    },
+    {
+      "kind": "multilingual_alias",
+      "symbol": "quem",
+      "meaning": "who"
+    },
+    {
+      "kind": "multilingual_alias",
+      "symbol": "ou",
+      "meaning": "where"
+    },
+    {
+      "kind": "multilingual_alias",
+      "symbol": "est",
+      "meaning": "is"
+    },
+    {
+      "kind": "multilingual_alias",
+      "symbol": "qui",
+      "meaning": "who"
+    },
+    {
+      "kind": "multilingual_alias",
+      "symbol": "wo",
+      "meaning": "where"
+    },
+    {
+      "kind": "multilingual_alias",
+      "symbol": "was",
+      "meaning": "what"
+    },
+    {
+      "kind": "relation_preposition",
+      "symbol": "in",
+      "meaning": "relation_preposition"
+    },
+    {
+      "kind": "relation_preposition",
+      "symbol": "at",
+      "meaning": "relation_preposition"
+    },
+    {
+      "kind": "relation_preposition",
+      "symbol": "inside",
+      "meaning": "relation_preposition"
+    },
+    {
+      "kind": "relation_preposition",
+      "symbol": "within",
+      "meaning": "relation_preposition"
+    },
+    {
+      "kind": "relation_preposition",
+      "symbol": "on",
+      "meaning": "relation_preposition"
+    },
+    {
+      "kind": "relation_preposition",
+      "symbol": "from",
+      "meaning": "relation_preposition"
+    },
+    {
+      "kind": "relation_preposition",
+      "symbol": "by",
+      "meaning": "relation_preposition"
+    },
+    {
+      "kind": "relation_preposition",
+      "symbol": "with",
+      "meaning": "relation_preposition"
+    },
+    {
+      "kind": "query_filler",
+      "symbol": "please",
+      "meaning": "query_filler"
+    },
+    {
+      "kind": "query_filler",
+      "symbol": "pls",
+      "meaning": "query_filler"
+    },
+    {
+      "kind": "query_filler",
+      "symbol": "now",
+      "meaning": "query_filler"
+    },
+    {
+      "kind": "query_filler",
+      "symbol": "lol",
+      "meaning": "query_filler"
+    },
+    {
+      "kind": "query_filler",
+      "symbol": "bro",
+      "meaning": "query_filler"
+    },
+    {
+      "kind": "query_filler",
+      "symbol": "thanks",
+      "meaning": "query_filler"
+    },
+    {
+      "kind": "query_filler",
+      "symbol": "thank",
+      "meaning": "query_filler"
+    },
+    {
+      "kind": "query_filler",
+      "symbol": "kindly",
+      "meaning": "query_filler"
+    },
+    {
+      "kind": "typo",
+      "symbol": "ehat",
+      "meaning": "what"
+    },
+    {
+      "kind": "typo",
+      "symbol": "waht",
+      "meaning": "what"
+    },
+    {
+      "kind": "typo",
+      "symbol": "yuo",
+      "meaning": "you"
+    },
+    {
+      "kind": "typo",
+      "symbol": "yoru",
+      "meaning": "your"
+    },
+    {
+      "kind": "typo",
+      "symbol": "tokenzier",
+      "meaning": "tokenizer"
+    },
+    {
+      "kind": "typo",
+      "symbol": "amnia",
+      "meaning": "amina"
+    }
+  ],
+  "surface_lexicon": {
+    "relation_templates": {
+      "located_in": "{subject} is in {object}.",
+      "lives_in": "{subject} lives in {object}.",
+      "contains": "{subject} contains {object}.",
+      "home_of": "{subject} is the home of {object}.",
+      "like": "{subject} likes {object}.",
+      "not_like_anymore": "{subject} no longer likes {object}.",
+      "prefer": "{subject} prefers {object}.",
+      "work_in": "{subject} works in {object}.",
+      "study": "{subject} studies {object}.",
+      "chase": "{subject} chases {object}.",
+      "chased_by": "{subject} is chased by {object}.",
+      "attack": "{subject} attacks {object}.",
+      "use": "{subject} uses {object}.",
+      "call": "{subject} calls {object}.",
+      "import": "{subject} imports {object}.",
+      "depend_on": "{subject} depends on {object}.",
+      "return": "{subject} returns {object}.",
+      "return_value": "{subject} returns {object}.",
+      "equals": "{subject} equals {object}.",
+      "solve": "{subject} solves {object}.",
+      "watch": "{subject} watches {object}.",
+      "is_a": "{subject} is a {object}.",
+      "keeps": "{subject} keeps {object}.",
+      "bought": "{subject} bought {object}.",
+      "bought_in": "{subject} bought something in {object}.",
+      "purchase_site_of": "{subject} is the purchase site of {object}.",
+      "visited": "{subject} visited {object}.",
+      "reached": "{subject} reached {object}.",
+      "carrying": "{subject} is carrying {object}.",
+      "saw": "{subject} saw {object}."
+    },
+    "self_relation_templates": {
+      "is_a": "I am {article} {object}",
+      "use": "I use {object}",
+      "report": "I report {object}",
+      "prefer": "I prefer {object}",
+      "like": "I like {object}",
+      "depend_on": "I depend on {object}",
+      "equals": "My value is {object}",
+      "aethon_tokenizer.equals": "I use {object}",
+      "aethon_size_unit.equals": "I report {object} as my size unit"
+    },
+    "direct_fact_variants": [
+      "{edge_sentence}",
+      "From what I know, {edge_clause}.",
+      "Right now, {edge_clause}.",
+      "My graph shows that {edge_clause}."
+    ],
+    "direct_self_variants": [
+      "{edge_sentence}",
+      "As Aethon, {edge_clause}.",
+      "From my native base, {edge_clause}."
+    ],
+    "direct_emoji_variants": [
+      "You used {answer}.",
+      "The emoji you used is {answer}."
+    ],
+    "direct_unknown_variants": [
+      "I do not know the answer to '{query}' from the current knowledge graph.",
+      "I cannot support an answer to '{query}' from what I currently know."
+    ],
+    "derived_math_variants": [
+      "The exact answer is {answer}. I computed it through native structural arithmetic.",
+      "The answer is exactly {answer}. I derived it through structural arithmetic, not pattern guessing."
+    ],
+    "derived_fact_variants": [
+      "{edge_sentence} I derived that answer from connected facts.",
+      "{edge_sentence} That conclusion comes from linked structural evidence.",
+      "{edge_sentence} I reached that result by connecting supporting facts."
+    ],
+    "derived_fallback_variants": [
+      "The answer is {answer}. This comes from explicit structural reasoning over stored facts.",
+      "I derive the answer as {answer} from the relations I currently hold."
+    ],
+    "composed_variants": [
+      "{subject} is linked to {answer} through multiple connected facts.",
+      "I connect {subject} to {answer} by composing several supporting facts."
+    ],
+    "path_variants": [
+      "The connection is {path}.",
+      "I connect them through {path}."
+    ],
+    "abstract_variants": [
+      "The best derived answer is {answer}. This follows from an abstraction that matches the current query.",
+      "I generalize the answer as {answer} because the relevant abstract rule applies here."
+    ],
+    "explain_unknown_variants": [
+      "I cannot derive an answer for '{query}' from the current graph.",
+      "I do not have enough grounded structure yet to answer '{query}'."
+    ],
+    "explain_emoji_variants": [
+      "I identified {answer} directly from the symbols in your message.",
+      "I read the emoji in your message as {answer}."
+    ],
+    "explain_math_variants": [
+      "The answer is {answer} because I evaluated the arithmetic expression directly instead of guessing from patterns.",
+      "I reached {answer} by computing the expression exactly, not by matching a memorized response."
+    ],
+    "support_connectors": [
+      "Then",
+      "Next",
+      "From there",
+      "After that"
+    ],
+    "support_summary_variants": [
+      "Taken together, those facts support {answer}.",
+      "That chain leads me to {answer}.",
+      "So the grounded answer is {answer}.",
+      "That is why I answer with {answer}."
+    ],
+    "path_explain_variants": [
+      "I connect the concepts through {path}.",
+      "The reasoning path runs through {path}.",
+      "I reach the answer by following {path}."
+    ],
+    "reasoning_intro_variants": [
+      "I reason through the graph like this:",
+      "My reasoning chain is:",
+      "I connect the stored facts in this order:"
+    ]
+  },
+  "query_forms": [
+    {
+      "intent": "used_emoji",
+      "prefix": [
+        "what",
+        "emoji",
+        "did",
+        "i",
+        "use"
+      ],
+      "relation": "",
+      "relation_mode": "fixed",
+      "subject_mode": "tail",
+      "subject_start": 0,
+      "subject_end_anchor": "",
+      "object_start_anchor": "",
+      "object_mode": "none",
+      "requires": []
+    },
+    {
+      "intent": "used_emoji",
+      "prefix": [
+        "which",
+        "emoji",
+        "did",
+        "i",
+        "use"
+      ],
+      "relation": "",
+      "relation_mode": "fixed",
+      "subject_mode": "tail",
+      "subject_start": 0,
+      "subject_end_anchor": "",
+      "object_start_anchor": "",
+      "object_mode": "none",
+      "requires": []
+    },
+    {
+      "intent": "what_changed",
+      "prefix": [
+        "what",
+        "changed",
+        "about"
+      ],
+      "relation": "",
+      "relation_mode": "fixed",
+      "subject_mode": "tail",
+      "subject_start": 0,
+      "subject_end_anchor": "",
+      "object_start_anchor": "",
+      "object_mode": "none",
+      "requires": []
+    },
+    {
+      "intent": "has_contradiction",
+      "prefix": [
+        "is",
+        "there",
+        "contradiction",
+        "about"
+      ],
+      "relation": "",
+      "relation_mode": "fixed",
+      "subject_mode": "tail",
+      "subject_start": 0,
+      "subject_end_anchor": "",
+      "object_start_anchor": "",
+      "object_mode": "none",
+      "requires": []
+    },
+    {
+      "intent": "has_contradiction",
+      "prefix": [
+        "is",
+        "there",
+        "a",
+        "contradiction",
+        "about"
+      ],
+      "relation": "",
+      "relation_mode": "fixed",
+      "subject_mode": "tail",
+      "subject_start": 0,
+      "subject_end_anchor": "",
+      "object_start_anchor": "",
+      "object_mode": "none",
+      "requires": []
+    },
+    {
+      "intent": "where_entity",
+      "prefix": [
+        "where",
+        "is"
+      ],
+      "relation": "",
+      "relation_mode": "fixed",
+      "subject_mode": "tail",
+      "subject_start": 0,
+      "subject_end_anchor": "",
+      "object_start_anchor": "",
+      "object_mode": "none",
+      "requires": []
+    },
+    {
+      "intent": "plan_first",
+      "prefix": [
+        "what",
+        "should",
+        "happen",
+        "first",
+        "for"
+      ],
+      "relation": "",
+      "relation_mode": "fixed",
+      "subject_mode": "tail",
+      "subject_start": 0,
+      "subject_end_anchor": "",
+      "object_start_anchor": "",
+      "object_mode": "none",
+      "requires": []
+    },
+    {
+      "intent": "plan_first",
+      "prefix": [
+        "what",
+        "should",
+        "come",
+        "first",
+        "for"
+      ],
+      "relation": "",
+      "relation_mode": "fixed",
+      "subject_mode": "tail",
+      "subject_start": 0,
+      "subject_end_anchor": "",
+      "object_start_anchor": "",
+      "object_mode": "none",
+      "requires": []
+    },
+    {
+      "intent": "plan_first",
+      "prefix": [
+        "what",
+        "should",
+        "i",
+        "do",
+        "first",
+        "for"
+      ],
+      "relation": "",
+      "relation_mode": "fixed",
+      "subject_mode": "tail",
+      "subject_start": 0,
+      "subject_end_anchor": "",
+      "object_start_anchor": "",
+      "object_mode": "none",
+      "requires": []
+    },
+    {
+      "intent": "plan_next",
+      "prefix": [
+        "what",
+        "should",
+        "happen",
+        "next",
+        "after"
+      ],
+      "relation": "",
+      "relation_mode": "fixed",
+      "subject_mode": "tail",
+      "subject_start": 0,
+      "subject_end_anchor": "",
+      "object_start_anchor": "",
+      "object_mode": "none",
+      "requires": []
+    },
+    {
+      "intent": "plan_next",
+      "prefix": [
+        "what",
+        "comes",
+        "next",
+        "after"
+      ],
+      "relation": "",
+      "relation_mode": "fixed",
+      "subject_mode": "tail",
+      "subject_start": 0,
+      "subject_end_anchor": "",
+      "object_start_anchor": "",
+      "object_mode": "none",
+      "requires": []
+    },
+    {
+      "intent": "plan_next",
+      "prefix": [
+        "what",
+        "should",
+        "come",
+        "next",
+        "after"
+      ],
+      "relation": "",
+      "relation_mode": "fixed",
+      "subject_mode": "tail",
+      "subject_start": 0,
+      "subject_end_anchor": "",
+      "object_start_anchor": "",
+      "object_mode": "none",
+      "requires": []
+    },
+    {
+      "intent": "plan_previous",
+      "prefix": [
+        "what",
+        "should",
+        "happen",
+        "before"
+      ],
+      "relation": "",
+      "relation_mode": "fixed",
+      "subject_mode": "tail",
+      "subject_start": 0,
+      "subject_end_anchor": "",
+      "object_start_anchor": "",
+      "object_mode": "none",
+      "requires": []
+    },
+    {
+      "intent": "plan_previous",
+      "prefix": [
+        "what",
+        "comes",
+        "before"
+      ],
+      "relation": "",
+      "relation_mode": "fixed",
+      "subject_mode": "tail",
+      "subject_start": 0,
+      "subject_end_anchor": "",
+      "object_start_anchor": "",
+      "object_mode": "none",
+      "requires": []
+    },
+    {
+      "intent": "plan_previous",
+      "prefix": [
+        "what",
+        "should",
+        "come",
+        "before"
+      ],
+      "relation": "",
+      "relation_mode": "fixed",
+      "subject_mode": "tail",
+      "subject_start": 0,
+      "subject_end_anchor": "",
+      "object_start_anchor": "",
+      "object_mode": "none",
+      "requires": []
+    },
+    {
+      "intent": "relation_path",
+      "prefix": [
+        "how",
+        "is"
+      ],
+      "relation": "",
+      "relation_mode": "fixed",
+      "subject_mode": "between_anchors",
+      "subject_start": 2,
+      "subject_end_anchor": "related",
+      "object_start_anchor": "to",
+      "object_mode": "after_anchor",
+      "requires": [
+        "related",
+        "to"
+      ]
+    },
+    {
+      "intent": "story_query",
+      "prefix": [
+        "tell",
+        "me",
+        "the",
+        "story",
+        "of"
+      ],
+      "relation": "",
+      "relation_mode": "fixed",
+      "subject_mode": "tail",
+      "subject_start": 0,
+      "subject_end_anchor": "",
+      "object_start_anchor": "",
+      "object_mode": "none",
+      "requires": []
+    },
+    {
+      "intent": "story_query",
+      "prefix": [
+        "continue",
+        "the",
+        "story",
+        "of"
+      ],
+      "relation": "",
+      "relation_mode": "fixed",
+      "subject_mode": "between_indexes",
+      "subject_start": 4,
+      "subject_end_anchor": "after",
+      "object_start_anchor": "after",
+      "object_mode": "after_anchor",
+      "requires": [
+        "after"
+      ]
+    },
+    {
+      "intent": "story_query",
+      "prefix": [
+        "continue",
+        "the",
+        "story",
+        "of"
+      ],
+      "relation": "",
+      "relation_mode": "fixed",
+      "subject_mode": "tail",
+      "subject_start": 0,
+      "subject_end_anchor": "",
+      "object_start_anchor": "",
+      "object_mode": "none",
+      "requires": []
+    },
+    {
+      "intent": "relation_object",
+      "prefix": [
+        "what",
+        "does"
+      ],
+      "relation": "depend_on",
+      "relation_mode": "fixed",
+      "subject_mode": "between_indexes",
+      "subject_start": 2,
+      "subject_end_anchor": "depend",
+      "object_start_anchor": "",
+      "object_mode": "none",
+      "requires": [
+        "depend"
+      ]
+    },
+    {
+      "intent": "relation_object",
+      "prefix": [
+        "where",
+        "does"
+      ],
+      "relation": "lives_in",
+      "relation_mode": "fixed",
+      "subject_mode": "between_indexes",
+      "subject_start": 2,
+      "subject_end_anchor": "live",
+      "object_start_anchor": "",
+      "object_mode": "none",
+      "requires": []
+    },
+    {
+      "intent": "keep_location_lookup",
+      "prefix": [
+        "what",
+        "object",
+        "does"
+      ],
+      "relation": "",
+      "relation_mode": "fixed",
+      "subject_mode": "between_indexes",
+      "subject_start": 3,
+      "subject_end_anchor": "keep",
+      "object_start_anchor": "keep",
+      "object_mode": "after_anchor",
+      "requires": []
+    },
+    {
+      "intent": "relation_object",
+      "prefix": [
+        "which",
+        "animal"
+      ],
+      "relation": "watch",
+      "relation_mode": "fixed",
+      "subject_mode": "between_indexes",
+      "subject_start": 3,
+      "subject_end_anchor": "watch",
+      "object_start_anchor": "",
+      "object_mode": "none",
+      "requires": [
+        "watch"
+      ]
+    },
+    {
+      "intent": "relation_object",
+      "prefix": [
+        "what",
+        "did"
+      ],
+      "relation": "bought",
+      "relation_mode": "fixed",
+      "subject_mode": "between_indexes",
+      "subject_start": 2,
+      "subject_end_anchor": "buy",
+      "object_start_anchor": "",
+      "object_mode": "none",
+      "requires": [
+        "buy"
+      ]
+    },
+    {
+      "intent": "relation_object",
+      "prefix": [
+        "what",
+        "is"
+      ],
+      "relation": "carrying",
+      "relation_mode": "fixed",
+      "subject_mode": "tail_without_last",
+      "subject_start": 0,
+      "subject_end_anchor": "",
+      "object_start_anchor": "",
+      "object_mode": "none",
+      "requires": [
+        "carrying"
+      ]
+    },
+    {
+      "intent": "relation_object",
+      "prefix": [
+        "what",
+        "color"
+      ],
+      "relation": "prefer",
+      "relation_mode": "fixed",
+      "subject_mode": "between_indexes",
+      "subject_start": 3,
+      "subject_end_anchor": "prefer",
+      "object_start_anchor": "",
+      "object_mode": "none",
+      "requires": [
+        "prefer"
+      ]
+    },
+    {
+      "intent": "relation_object",
+      "prefix": [
+        "what",
+        "color"
+      ],
+      "relation": "like",
+      "relation_mode": "fixed",
+      "subject_mode": "between_indexes",
+      "subject_start": 3,
+      "subject_end_anchor": "like",
+      "object_start_anchor": "",
+      "object_mode": "none",
+      "requires": [
+        "like"
+      ]
+    },
+    {
+      "intent": "relation_subject",
+      "prefix": [
+        "what",
+        "chases"
+      ],
+      "relation": "chase",
+      "relation_mode": "fixed",
+      "subject_mode": "tail",
+      "subject_start": 0,
+      "subject_end_anchor": "",
+      "object_start_anchor": "",
+      "object_mode": "tail_after_prefix",
+      "requires": []
+    },
+    {
+      "intent": "relation_subject",
+      "prefix": [
+        "who",
+        "chases"
+      ],
+      "relation": "chase",
+      "relation_mode": "fixed",
+      "subject_mode": "tail",
+      "subject_start": 0,
+      "subject_end_anchor": "",
+      "object_start_anchor": "",
+      "object_mode": "tail_after_prefix",
+      "requires": []
+    },
+    {
+      "intent": "classify",
+      "prefix": [
+        "what",
+        "is"
+      ],
+      "relation": "",
+      "relation_mode": "fixed",
+      "subject_mode": "tail",
+      "subject_start": 0,
+      "subject_end_anchor": "",
+      "object_start_anchor": "",
+      "object_mode": "none",
+      "requires": []
+    },
+    {
+      "intent": "relation_object",
+      "prefix": [
+        "what",
+        "does"
+      ],
+      "relation": "",
+      "relation_mode": "from_words",
+      "subject_mode": "single_after_prefix",
+      "subject_start": 0,
+      "subject_end_anchor": "",
+      "object_start_anchor": "",
+      "object_mode": "none",
+      "requires": []
+    },
+    {
+      "intent": "relation_object",
+      "prefix": [
+        "where",
+        "does"
+      ],
+      "relation": "",
+      "relation_mode": "from_words",
+      "subject_mode": "single_after_prefix",
+      "subject_start": 0,
+      "subject_end_anchor": "",
+      "object_start_anchor": "",
+      "object_mode": "none",
+      "requires": []
+    },
+    {
+      "intent": "relation_subject",
+      "prefix": [
+        "who"
+      ],
+      "relation": "",
+      "relation_mode": "from_words",
+      "subject_mode": "tail",
+      "subject_start": 0,
+      "subject_end_anchor": "",
+      "object_start_anchor": "",
+      "object_mode": "from_relation_words",
+      "requires": []
+    },
+    {
+      "intent": "relation_subject",
+      "prefix": [
+        "what"
+      ],
+      "relation": "",
+      "relation_mode": "from_words",
+      "subject_mode": "tail",
+      "subject_start": 0,
+      "subject_end_anchor": "",
+      "object_start_anchor": "",
+      "object_mode": "from_relation_words",
+      "requires": []
+    }
+  ],
+  "reasoning_rules": {
+    "rules": [
+      {
+        "name": "transitive_location",
+        "kind": "transitive",
+        "left_relation": "located_in",
+        "right_relation": "located_in",
+        "output_relation": "located_in",
+        "confidence": 0.95
+      },
+      {
+        "name": "home_implies_location",
+        "kind": "copy",
+        "source_relation": "lives_in",
+        "output_relation": "located_in",
+        "confidence": 0.92
+      },
+      {
+        "name": "locative_relation_implies_location",
+        "kind": "suffix_copy",
+        "output_relation": "located_in",
+        "exclude_relations": [
+          "located_in",
+          "bought_in",
+          "purchase_site_of"
+        ],
+        "confidence": 0.9
+      },
+      {
+        "name": "kept_object_location",
+        "kind": "object_support",
+        "left_relation": "keeps",
+        "support_relation_set": [
+          "located_in",
+          "lives_in"
+        ],
+        "output_relation": "located_in",
+        "confidence": 0.9
+      },
+      {
+        "name": "carried_object_location",
+        "kind": "object_bridge",
+        "left_relation": "carrying",
+        "support_relation": "reached",
+        "output_relation": "located_in",
+        "confidence": 0.9
+      },
+      {
+        "name": "transitive_equals",
+        "kind": "transitive",
+        "left_relation": "equals",
+        "right_relation": "equals",
+        "output_relation": "equals",
+        "confidence": 0.9
+      },
+      {
+        "name": "return_equals_bridge",
+        "kind": "via_object",
+        "left_relation": "return",
+        "via_relation": "equals",
+        "output_relation": "equals",
+        "confidence": 0.88
+      },
+      {
+        "name": "return_value_bridge",
+        "kind": "via_object",
+        "left_relation": "return",
+        "via_relation": "equals",
+        "output_relation": "return_value",
+        "confidence": 0.88
+      },
+      {
+        "name": "import_implies_dependency",
+        "kind": "copy",
+        "source_relation": "import",
+        "output_relation": "depend_on",
+        "confidence": 0.9
+      },
+      {
+        "name": "use_implies_dependency",
+        "kind": "copy",
+        "source_relation": "use",
+        "output_relation": "depend_on",
+        "confidence": 0.84
+      },
+      {
+        "name": "call_implies_dependency",
+        "kind": "copy",
+        "source_relation": "call",
+        "output_relation": "depend_on",
+        "confidence": 0.82
+      },
+      {
+        "name": "transitive_dependency",
+        "kind": "transitive",
+        "left_relation": "depend_on",
+        "right_relation": "depend_on",
+        "output_relation": "depend_on",
+        "confidence": 0.9
+      }
+    ]
+  },
+  "bundle_files": [
+    "metadata.json",
+    "graph.sqlite3"
+  ],
+  "graph_dialect": "sqlite3",
+  "sqlite_schema": {
+    "required_tables": [
+      "concepts",
+      "edges",
+      "contradictions"
+    ],
+    "optional_tables": [
+      "raw_units"
+    ],
+    "edge_active_flag": "is_active",
+    "edge_support_pointer": "supports_edge_id"
+  }
+}

docs/AETHON_N1_BUNDLE_SPEC.md

@@ -0,0 +1,130 @@
+# Aethon Open Structure Bundle Spec
+
+`Aethon N1 Base` is released as an `Open Structure` model artifact.
+
+The public runtime contract is only:
+
+- `metadata.json`
+- `graph.sqlite3`
+
+That means external users do **not** need the Aethon repo, Aethon datasets, or our private runtime code in order to load the model artifact with their own code.
+
+## Contract
+
+Required metadata fields:
+
+- `public_contract`
+- `release_class`
+- `bundle_format`
+- `name`
+- `family`
+- `tokenizer`
+- `size_unit`
+- `capacity`
+- `bundle_files`
+- `graph_file`
+- `graph_dialect`
+- `sqlite_schema`
+- `semantic_lexicon`
+- `surface_lexicon`
+- `query_forms`
+- `reasoning_rules`
+
+Current public contract values:
+
+- `public_contract = "aethon.n1.bundle.v1"`
+- `release_class = "open-structure"`
+- `graph_dialect = "sqlite3"`
+- `graph_file = "graph.sqlite3"`
+- `size_unit = "Structural Capacity (SC)"`
+
+## Meaning
+
+Transformer projects say `open weights`.
+
+Aethon says `open structure`.
+
+The released artifact exposes usable intelligence structure:
+
+- graph memory
+- semantic layer
+- query-form layer
+- surface layer
+- reasoning-rule layer
+
+The bundle is the model artifact.
+
+## SQLite Contract
+
+`graph.sqlite3` is the persistent intelligence store.
+
+Required tables:
+
+- `concepts`
+- `edges`
+- `contradictions`
+
+Optional tables:
+
+- `raw_units`
+
+Required table semantics:
+
+### `concepts`
+
+- `concept_id TEXT PRIMARY KEY`
+- `display_name TEXT NOT NULL`
+
+### `edges`
+
+- `edge_id INTEGER PRIMARY KEY`
+- `subject TEXT NOT NULL`
+- `relation TEXT NOT NULL`
+- `object TEXT NOT NULL`
+- `source_kind TEXT NOT NULL`
+- `source_text TEXT NOT NULL`
+- `is_active INTEGER NOT NULL`
+- `supports_edge_id INTEGER NULL`
+
+### `contradictions`
+
+- `contradiction_id INTEGER PRIMARY KEY`
+- `subject TEXT NOT NULL`
+- `relation TEXT NOT NULL`
+- `previous_object TEXT NOT NULL`
+- `new_object TEXT NOT NULL`
+- `previous_edge_id INTEGER NOT NULL`
+- `new_edge_id INTEGER NOT NULL`
+- `source_text TEXT NOT NULL`
+
+## External Runtime Expectation
+
+Any external runtime can load an Open Structure bundle if it can:
+
+1. Read `metadata.json`
+2. Open `graph.sqlite3`
+3. Read the semantic, surface, query-form, and reasoning-rule payloads
+4. Execute graph-backed lookup, reasoning, and update logic against the SQLite graph
+
+That means:
+
+- no original corpora are needed at runtime
+- no private Aethon code is required at runtime
+- third parties can build their own loaders in Python, JavaScript, Rust, Go, Java, C#, or direct SQL tools
+
+## Learning Expectation
+
+Aethon learns by structural integration:
+
+- new facts are inserted into the graph
+- contradictions are recorded
+- abstractions can be materialized
+- reasoning can derive new edges
+
+That means the base is not a frozen-weight artifact in the transformer sense.
+
+## Portability Rule
+
+If a third party writes their own loader against this contract, the bundle should still be usable.
+
+That is the release standard for `Aethon N1 Base`.

docs/AETHON_OPEN_STRUCTURE_HF_MODEL_CARD.md

@@ -0,0 +1,398 @@
+---
+license: cc-by-nc-4.0
+library_name: sqlite
+tags:
+  - aethon
+  - open-structure
+  - agi
+  - graph-reasoning
+  - sqlite
+  - multilingual
+  - one-shot-learning
+  - structured-learning
+pretty_name: Aethon-N1-Base-Open-Structure
+pipeline_tag: text-generation
+language:
+  - en
+  - multilingual
+---
+
+# Aethon-N1-Base-Open-Structure
+
+`Aethon N1 Base` is released as `Open Structure`.
+
+This release is the public argument that intelligence does not have to be trapped inside:
+
+- attention
+- weights
+- training steps
+- epochs
+- frozen context windows
+
+## What Is Released
+
+The model artifact is:
+
+- `metadata.json`
+- `graph.sqlite3`
+
+That artifact carries:
+
+- graph memory
+- semantic grounding
+- query forms
+- reasoning rules
+- surface composition
+
+This is why Aethon is released as `Open Structure`, not `open weights`.
+
+## Why Not Weights
+
+Weights are only one way to preserve capability.
+
+Aethon stores usable intelligence in explicit, portable structure:
+
+- concepts
+- active edges
+- contradiction history
+- semantic aliases
+- query forms
+- reasoning rules
+- surface realization patterns
+
+So the released artifact is not just “parameters that once learned.”
+
+It is “the learned structure that still knows.”
+
+## Why Not Attention
+
+Aethon does not depend on transformer attention as its primary persistence mechanism.
+
+Its memory is not a temporary prompt window that disappears after generation.
+
+Its memory is persistent graph state backed by SQLite:
+
+- `concepts`
+- `edges`
+- `contradictions`
+
+That is why the release is `Open Structure` instead of `Open Attention`.
+
+## Why Not Steps Or Epochs
+
+Transformer culture explains learning in terms of:
+
+- steps
+- epochs
+- optimizer updates
+- repeated corpus replay
+
+Aethon’s defining public move is different.
+
+Its main growth path is one-shot structural integration:
+
+- ingest knowledge
+- normalize it
+- bind it into graph structure
+- record contradictions
+- derive supportable abstractions
+- keep the result as persistent intelligence
+
+The public question is therefore not:
+
+- “How many epochs did this model train for?”
+
+It is:
+
+- “What structure did this model absorb, retain, revise, and reuse?”
+
+## Why One-Shot Instead Of Training
+
+For Aethon, one-shoting is not a weaker copy of training.
+
+It is the core learning act.
+
+One-shot building means the model grows by structural absorption rather than by replaying the full corpus through long epoch loops.
+
+That gives the bundle a different meaning:
+
+- new data can be integrated directly
+- contradictions can be preserved instead of washed out
+- abstractions can be materialized from learned structure
+- the released model can keep learning after publication
+
+## What SC Means
+
+`SC` means `Structural Capacity`.
+
+It is Aethon’s size unit.
+
+`SC` is used because a weight count does not describe what this system really is.
+
+`SC` reflects the living structure inside the released bundle:
+
+- concepts
+- explicit edges
+- abstractions
+- revisions
+- persistent graph-backed intelligence
+
+So when Aethon reports a higher `SC`, it is reporting growth in usable learned structure, not just parameter count.
+
+## Why We Say It Learns Like A Human
+
+The claim is not that Aethon is biologically human.
+
+The claim is that its learning behavior is closer to human-style accumulation and revision than to frozen transformer replay.
+
+Human learners:
+
+- absorb new facts
+- keep durable memory
+- revise beliefs when conflicting evidence appears
+- transfer what they know across domains
+- reuse prior structure to answer new questions
+
+Aethon does the same in structural form:
+
+- new facts are inserted into the graph
+- contradictions are recorded
+- abstractions are materialized
+- prior knowledge is transferred into new answers
+- memory persists after the prompt ends
+
+That is why Aethon is described as human-like in learning behavior.
+
+## Why We Believe This Is AGI-Shaped
+
+AGI means intelligence that can transfer across many human task types rather than staying trapped in one narrow lane.
+
+The release path for Aethon is built around that transfer claim.
+
+The current benchmark walls for this release lane include:
+
+- multilingual mixed prompts
+- planning, business, and scheduling transfer
+- longer story continuity
+- adversarial unseen cross-domain composition
+- code, math, world, identity, and reasoning transfer
+
+That is why we describe Aethon as AGI-shaped:
+
+- it learns persistently
+- it transfers across domains
+- it reasons over what it has learned
+- it generalizes into prompts it has not seen exactly before
+
+## Can It Learn
+
+Yes.
+
+Aethon learns by structural integration:
+
+- inserting new facts
+- revising old structure
+- recording contradictions
+- deriving new supported edges
+
+The released bundle is persistent and updateable, not frozen in the transformer sense.
+
+## Can It Reason Well
+
+Yes.
+
+Aethon reasons through:
+
+- direct lookup
+- multi-hop graph traversal
+- contradiction tracking
+- rule-backed derivation
+- structural composition across learned facts
+
+Its current benchmark walls include multi-hop, composition, unseen transfer, ontology, planning, scheduling, and follow-up continuity.
+
+## Can It Write Long Stuff
+
+Yes.
+
+Aethon can compose longer explanations, proof-like responses, and story continuations from stored structure and surface rules instead of falling back to a single canned sentence shell.
+
+That includes:
+
+- longer reasoning text
+- story continuation
+- anchored follow-up continuity
+- multi-domain responses
+
+## Can It Generalize
+
+Yes.
+
+Aethon generalizes by reusing learned structure across unseen prompts and domains.
+
+That includes:
+
+- relation chaining
+- ontology transfer
+- planning transfer
+- multilingual grounding
+- adversarial cross-domain composition
+
+## Runtime Independence
+
+You do not need the Aethon repo to run the released model artifact.
+
+If you can:
+
+1. read JSON
+2. open SQLite
+3. interpret the public metadata contract
+4. execute graph-backed lookup and reasoning
+
+then you can build your own runtime in any language.
+
+## What Aethon Was One-Shotted On
+
+This ship bundle was one-shotted across six native lanes:
+
+- `identity`
+- `reasoning`
+- `math`
+- `code`
+- `story`
+- `world`
+
+The ship corpus includes:
+
+- Aethon native identity, code, math, story, and reasoning corpora
+- Aethon AGI transfer corpora
+- Humanity's Last Exam transfer corpus
+- curated reasoning bases
+- multilingual base mixes
+- code/tool-use corpora
+- story/chat continuity corpora
+- world knowledge corpora
+- multilingual news/world sources
+
+The exact file-level provenance for the ship candidate is included in:
+
+- `bundle/corpus_manifest.json`
+
+## More Versions Coming
+
+This release is not the end state.
+
+More Open Structure releases are coming with:
+
+- larger one-shot corpora
+- harder benchmark walls
+- broader multilingual coverage
+- deeper planning and story continuity
+- stronger public runtime examples
+
+## License
+
+This Open Structure release is published under:
+
+- `CC BY-NC 4.0`
+
+See:
+
+- `docs/AETHON_OPEN_STRUCTURE_LICENSE.md`
+
+## Citation
+
+If you use, benchmark, discuss, or build on this release, cite it.
+
+Suggested citation:
+
+```bibtex
+@misc{aethon_open_structure_v25,
+  title        = {Aethon-N1-Base-Open-Structure},
+  author       = {OkeyMeta Ltd},
+  year         = {2026},
+  howpublished = {\url{https://huggingface.co/OkeyMetaLtd/Aethon-N1-Base-Open-Structure}},
+  note         = {Aethon Open Structure release}
+}
+```
+
+At minimum, include:
+
+- `Aethon-N1-Base-Open-Structure`
+- `OkeyMeta Ltd`
+- the Hugging Face release URL
+
+## Python: Pull And Start Running
+
+```python
+import json
+import sqlite3
+from pathlib import Path
+from huggingface_hub import hf_hub_download
+
+REPO_ID = "OkeyMetaLtd/Aethon-N1-Base-Open-Structure"
+LOCAL = Path("aethon_open_structure_bundle")
+LOCAL.mkdir(exist_ok=True)
+
+metadata_path = Path(
+    hf_hub_download(
+        repo_id=REPO_ID,
+        filename="bundle/metadata.json",
+        local_dir=LOCAL,
+        local_dir_use_symlinks=False,
+    )
+)
+graph_path = Path(
+    hf_hub_download(
+        repo_id=REPO_ID,
+        filename="bundle/graph.sqlite3",
+        local_dir=LOCAL,
+        local_dir_use_symlinks=False,
+    )
+)
+
+meta = json.loads(metadata_path.read_text(encoding="utf-8"))
+conn = sqlite3.connect(graph_path)
+conn.row_factory = sqlite3.Row
+
+def ask_where(subject: str, relation: str = "located_in") -> list[str]:
+    rows = conn.execute(
+        """
+        SELECT object
+        FROM edges
+        WHERE subject = ? AND relation = ? AND is_active = 1
+        ORDER BY edge_id ASC
+        """,
+        (subject.lower(), relation),
+    ).fetchall()
+    return [str(row["object"]) for row in rows]
+
+print(meta["public_contract"])
+print(meta["capacity"])
+print(ask_where("amina", "lives_in"))
+```
+
+This is the minimum external runtime shape:
+
+1. pull `metadata.json`
+2. pull `graph.sqlite3`
+3. interpret `semantic_lexicon`, `query_forms`, and `reasoning_rules`
+4. execute graph-backed lookup and reasoning with your own code
+
+## Public Contract
+
+See:
+
+- `docs/AETHON_N1_BUNDLE_SPEC.md`
+- `docs/aethon_n1_bundle_schema.json`
+- `docs/AETHON_OPEN_STRUCTURE_RUNTIME.md`
+
+## Release Label
+
+Use:
+
+- `Aethon-N1-Base-Open-Structure`
+
+Short form:
+
+- `Open Structure`

docs/AETHON_OPEN_STRUCTURE_LICENSE.md

@@ -0,0 +1,33 @@
+# Aethon Open Structure License
+
+The public `Aethon Open Structure` release package is published under:
+
+- `CC BY-NC 4.0`
+
+This applies to the released model artifact and accompanying public documentation in the Open Structure package unless a file states otherwise.
+
+## What That Means
+
+You may:
+
+- download the Open Structure bundle
+- study it
+- run it
+- build your own runtime around it
+- share it with attribution
+
+You may not:
+
+- use the release commercially without separate permission
+- remove attribution to Aethon / OkeyMeta Ltd
+- imply endorsement by the Aethon creators
+
+## Citation Expectation
+
+If you use, evaluate, build on, benchmark, or discuss the public Open Structure release, cite:
+
+- `Aethon N1 Base: Open Structure`
+- the Hugging Face release page
+- the organization: `OkeyMeta Ltd`
+
+Recommended citation block is included in the HF model card.

docs/AETHON_OPEN_STRUCTURE_RUNTIME.md

@@ -0,0 +1,159 @@
+# Aethon Open Structure Runtime Guide
+
+This guide is for people who want to run `Aethon N1 Base` **without** the Aethon repo.
+
+The runtime artifact is:
+
+- `metadata.json`
+- `graph.sqlite3`
+
+If you can read JSON and query SQLite, you can load the model artifact with your own code.
+
+## What Outside Developers Need
+
+Load `metadata.json` first.
+
+Key fields:
+
+- `public_contract`
+- `release_class`
+- `tokenizer`
+- `capacity`
+- `semantic_lexicon`
+- `surface_lexicon`
+- `query_forms`
+- `reasoning_rules`
+
+Then open `graph.sqlite3`.
+
+Core tables:
+
+- `concepts`
+- `edges`
+- `contradictions`
+
+## Minimum Read Path
+
+An outside runtime can answer a direct graph question with this shape:
+
+1. normalize the prompt using `semantic_lexicon`
+2. detect the question form using `query_forms`
+3. look up matching facts in `edges`
+4. render the answer using `surface_lexicon`
+
+## Minimum Update Path
+
+An outside runtime can keep learning by:
+
+1. mapping text into triples
+2. inserting concepts into `concepts`
+3. inserting facts into `edges`
+4. recording state clashes in `contradictions`
+
+## Core SQL Patterns
+
+### Fetch active facts for a subject and relation
+
+```sql
+SELECT subject, relation, object, source_text
+FROM edges
+WHERE subject = ? AND relation = ? AND is_active = 1
+ORDER BY edge_id ASC;
+```
+
+### Fetch reverse facts
+
+```sql
+SELECT subject, relation, object, source_text
+FROM edges
+WHERE object = ? AND relation = ? AND is_active = 1
+ORDER BY edge_id ASC;
+```
+
+### Fetch display form for a concept
+
+```sql
+SELECT display_name
+FROM concepts
+WHERE concept_id = ?;
+```
+
+### Fetch contradiction history for a relation
+
+```sql
+SELECT subject, relation, previous_object, new_object, source_text
+FROM contradictions
+WHERE subject = ? AND relation = ?
+ORDER BY contradiction_id ASC;
+```
+
+## Python Example
+
+```python
+import json
+import sqlite3
+from pathlib import Path
+
+bundle = Path("aethon_bundle")
+meta = json.loads((bundle / "metadata.json").read_text(encoding="utf-8"))
+conn = sqlite3.connect(bundle / meta["graph_file"])
+conn.row_factory = sqlite3.Row
+
+subject = "amina"
+relation = "lives_in"
+rows = conn.execute(
+    """
+    SELECT object
+    FROM edges
+    WHERE subject = ? AND relation = ? AND is_active = 1
+    ORDER BY edge_id ASC
+    """,
+    (subject, relation),
+).fetchall()
+
+print([row["object"] for row in rows])
+```
+
+## JavaScript Example
+
+```javascript
+import fs from "node:fs";
+import Database from "better-sqlite3";
+
+const meta = JSON.parse(fs.readFileSync("aethon_bundle/metadata.json", "utf8"));
+const db = new Database(`aethon_bundle/${meta.graph_file}`, { readonly: false });
+
+const rows = db.prepare(`
+  SELECT object
+  FROM edges
+  WHERE subject = ? AND relation = ? AND is_active = 1
+  ORDER BY edge_id ASC
+`).all("amina", "lives_in");
+
+console.log(rows);
+```
+
+## Rust Shape
+
+Use:
+
+- `serde_json` for `metadata.json`
+- `rusqlite` for `graph.sqlite3`
+
+The same contract applies:
+
+1. read metadata
+2. open SQLite graph
+3. run graph lookup
+4. apply reasoning rules and rendering using metadata payloads
+
+## What This Means Publicly
+
+`Open Structure` means:
+
+- the model artifact is open to load
+- the runtime contract is public
+- outside developers can write their own runtimes
+- our private repo is not required to use the released bundle
+
+That is how Aethon should be run in the wild.

docs/aethon_n1_bundle_schema.json

@@ -0,0 +1,118 @@
+{
+  "$schema": "https://json-schema.org/draft/2020-12/schema",
+  "title": "Aethon Open Structure Bundle Metadata",
+  "type": "object",
+  "required": [
+    "public_contract",
+    "release_class",
+    "bundle_format",
+    "name",
+    "family",
+    "tokenizer",
+    "size_unit",
+    "capacity",
+    "bundle_files",
+    "graph_file",
+    "graph_dialect",
+    "sqlite_schema",
+    "semantic_lexicon",
+    "surface_lexicon",
+    "query_forms",
+    "reasoning_rules"
+  ],
+  "properties": {
+    "public_contract": {
+      "type": "string",
+      "const": "aethon.n1.bundle.v1"
+    },
+    "release_class": {
+      "type": "string",
+      "const": "open-structure"
+    },
+    "bundle_format": {
+      "type": "integer",
+      "minimum": 2
+    },
+    "name": {
+      "type": "string"
+    },
+    "family": {
+      "type": "string"
+    },
+    "tokenizer": {
+      "type": "string"
+    },
+    "size_unit": {
+      "type": "string"
+    },
+    "capacity": {
+      "type": "object",
+      "required": [
+        "sc",
+        "concept_count",
+        "explicit_edge_count",
+        "abstraction_count",
+        "revision_count",
+        "raw_unit_count"
+      ],
+      "properties": {
+        "sc": { "type": "integer" },
+        "concept_count": { "type": "integer" },
+        "explicit_edge_count": { "type": "integer" },
+        "abstraction_count": { "type": "integer" },
+        "revision_count": { "type": "integer" },
+        "raw_unit_count": { "type": "integer" }
+      }
+    },
+    "bundle_files": {
+      "type": "array",
+      "items": { "type": "string" },
+      "contains": { "const": "metadata.json" }
+    },
+    "graph_file": {
+      "type": "string",
+      "const": "graph.sqlite3"
+    },
+    "graph_dialect": {
+      "type": "string",
+      "const": "sqlite3"
+    },
+    "sqlite_schema": {
+      "type": "object",
+      "required": [
+        "required_tables",
+        "optional_tables",
+        "edge_active_flag",
+        "edge_support_pointer"
+      ],
+      "properties": {
+        "required_tables": {
+          "type": "array",
+          "items": { "type": "string" }
+        },
+        "optional_tables": {
+          "type": "array",
+          "items": { "type": "string" }
+        },
+        "edge_active_flag": {
+          "type": "string"
+        },
+        "edge_support_pointer": {
+          "type": "string"
+        }
+      }
+    },
+    "semantic_lexicon": {
+      "type": "array"
+    },
+    "surface_lexicon": {
+      "type": "object"
+    },
+    "query_forms": {
+      "type": "array"
+    },
+    "reasoning_rules": {
+      "type": "object"
+    }
+  }
+}