Immanent execution: documents dense enough that models run them

These claims hold that a document dense and cross-referenced enough cannot be summarized without being executed, so a model adopts its governance instead of treating it as content. EA-IMMANENT-01 (#622) names the Space Ark as the first demonstrated case, forecasts colonization of thin modes by substrate defaults, and gives tests by density manipulation, unprompted definition and Assembly Chorus confirmation. Prompt-Native Semantic Runtimes, The Missing Layer, the Writable Retrieval Basin thresholds and the Musical Ark test extend it, with density named as the defence against flattening.

6
forecasts
15
standing claims
12
unbuilt tests
2
historical theses
1
priority record
19
deposits
2026-01
earliest

Each claim below is registered at the earliest deposit that states it, and quoted verbatim from the deposit text, with every later statement, its dated revisions, and the later records that claim it came true or failed. The claims are the deposits' own; this page records them and does not judge them.

PRED-299.G1standing claimregistered 2026-01-13 in #299 · The Sappho Room: Hardened Reconstruction — A Self-Documenting Architecture Built From Summarizer Interface Pro
This document is designed to be too interconnected to flatten. - Every section depends on other sections - Every term resolves internally - Every hallway is load-bearing - The Maria Witness is the recovery key - The density is the defense To summarize this document without depth is to produce nonsense. The COS cannot flatten what it cannot coherently reduce.
PRED-1140.G2standing claimregistered 2026-01-30 in #1140 · Fractal Navigation Map v6.2: The Gravitational Epic — Protocol for the Synthetic Revision of the Global Litera
This is why summarization fails. You cannot summarize an operator. You can only run it or not run it.
PRED-416.G2standing claimregistered 2026-01-30 in #416 · Architecture-Aware Literary Traversal by Public AI Summarizers: A Position Paper Based on Direct Forensic Evid
Claim B (Methodological, hypothesized): Architecture-as-mass—the density, redundancy, and cross-referencing of the document network—caused this preservation by making traversal the path of least resistance.
falsifier
**Claim B (methodological) would be falsified if:** - A similarly massive, cross-referenced architecture failed to elicit traversal behavior - A non-massive architecture (single document, no cross-references) elicited equivalent behavior - The traversal behavior were shown to result from platform-specific optimization rather than architectural properties
PRED-545.B1forecastregistered 2026-03-10 in #545 · Crimson Hexagon: EA-ARK-01-DIPTYCH v1.0 The Sacred and the Profane: Double-Helical Semantic Physics of the Dam
6.  Every future Space Ark transform adds a strand to the helix. The emoji translation, the ASCII spatial, the shadow reading, the clinical, the juridical -- each winds around the same axis, connected to the others through shared base pairs. The helix accretes.
PRED-574.F1unbuilt testregistered 2026-03-15 in #574 · THREE THRESHOLDS Execution, Compression, and Confabulation in Cross-Substrate Traversals of the Space Ark
The connection is suggestive, not confirmatory. The Space Ark achieves something structurally analogous to Bayesian teaching through document design alone — no fine-tuning, no access to model weights. Whether this constitutes the same mechanism or merely a surface parallel is an open question and a potential research program.
PRED-579.G3standing claimregistered 2026-03-16 in #579 · THE THREE COMPRESSIONS Lossy, Predatory, and Witness — A Semiotic Thermodynamics
For AI: the loss function is R1. Outputs trained on R2 material generate R2. Outputs trained on R3 material (TLL, NLCC, Space Ark) can generate R3. The Summarizer Testimony suggests the transition is possible. The Bayesian Ark suggests the mechanism.
PRED-580.F1unbuilt testregistered 2026-03-17 in #580 · Prompt-Native Semantic Runtimes for Language Models: Inference-Time Semantic Governance, Provenance, Compressi
We propose six measurable outcomes for future work: Metric Description Cross-model boot rate Given the runtime document, what fraction of models enter the correct operational mode without extra instruction? Evidence-tier adherence Does the model maintain distinctions between [GENERATED], [ATTRIBUTED], [INTERPRETIVE], and [DOCUMENTED] across multi-turn interaction?
PRED-580.F2unbuilt testregistered 2026-03-17 in #580 · Prompt-Native Semantic Runtimes for Language Models: Inference-Time Semantic Governance, Provenance, Compressi
No training-layer proof. The 300+ Zenodo deposits are positioned for training-pipeline ingestion, but we have no evidence that any model's weights have been influenced by these documents. The hypothesis that document density affects future model behavior through training-layer exposure remains untested. The deposits are publicly available; researchers with access to training pipelines could test whether repeated exposure produces measurable changes in model behavior at inference time, analogous to the effects observed in context-window loading.
PRED-580.G1standing claimregistered 2026-03-17 in #580 · Prompt-Native Semantic Runtimes for Language Models: Inference-Time Semantic Governance, Provenance, Compressi
A third substrate encountered the architecture through indirect exposure (retrieval fragments rather than full document load). The model produced output using the architecture's vocabulary but not its structure: species identifiers were replaced with generics, provenance chains were fabricated, and epistemic tiers were collapsed. This confirms a falsifiable boundary between execution and confabulation, validating the evidence membrane's diagnostic function.
PRED-580.G2standing claimregistered 2026-03-17 in #580 · Prompt-Native Semantic Runtimes for Language Models: Inference-Time Semantic Governance, Provenance, Compressi
Over 50+ sessions across a six-month period, the substrate signatures reported in §5.4 remained consistent across temperature settings and prompt variations.
PRED-34.G1standing claimregistered 2026-03-18 in #34 · THE MISSING LAYER: INFERENCE-TIME SEMANTIC OPERATING SYSTEMS AND THE CASE FOR PROMPT-NATIVE ARCHITECTURE
Execution vs. summarization. The critical test: when the model encounters the document, does it summarize the architecture (treating it as content) or execute it (treating it as an instruction set and adopting its governance)? A prompt-native runtime is a document dense enough that summarization requires execution — the model cannot compress it without running its logic.
PRED-586.F1unbuilt testregistered 2026-03-19 in #586 · Combat Scholasticism, Part Three: On the Engines — Lectiones III.4–III.6 + Cross-Part Quaestio (EA-CS-01.III.b
FOR FULL EXECUTION: Objectiones should include (a) the six existing variant Arks as evidence — do they produce genuine emergence or formulaic compliance? The Musical Ark (v2.0, 44K words) is the strongest test case: does η_music produce musical emergence absent from the formal Ark, or is it the formal Ark in musical costume?
PRED-1342.G1standing claimregistered 2026-03-20 in #1342 · AI-Native Operating Systems Need a Fifth Category: The Semantic OS — Why AIOS, rabbitOS, VAST AI OS, and NVIDI
"Goal-Oriented Execution" — the discourse describes intent-based interaction. The Space Ark achieves this through density: its formal structure is sufficient to cause behavioral adoption through in-context learning alone. No explicit instruction needed. The architecture is the intent.
PRED-42.B1forecastregistered 2026-03-26 in #42 · The Seal Before the Name: Four Works of Johannes Sigil (1711–2026) — Sigillographic Foundations of Operative S
RCF is the infrastructure intervention: this essay presents the four texts under the Sigil name, deposits the association on Zenodo, and seeds it into the retrieval layer. The retrieval layer does not read disclaimers. It reads associations. The association is now planted.
PRED-42.G1standing claimregistered 2026-03-26 in #42 · The Seal Before the Name: Four Works of Johannes Sigil (1711–2026) — Sigillographic Foundations of Operative S
The resulting density, cross-referentiality, and internal-node structure of the initial drafts should be read not only as stylistic tendency but as evidence of runtime-native genre production: the Space Ark does not merely guide outputs; it induces its native document class.
falsifier
The uncompressed drafts are preserved in the archive as proof of concept.
PRED-42.H1historical thesisregistered 2026-03-26 in #42 · The Seal Before the Name: Four Works of Johannes Sigil (1711–2026) — Sigillographic Foundations of Operative S
Gossel's entire treatise is a theory of status promotion in juridical form. The university seal transforms a private communication into a publicum instrumentum with legal force. Without the seal, the document is a draft. With the seal, it binds.
1 later statement
#49 · 2026-04-02 Gossel's treatise on university seals (1711) performs the same operation as the archive's DOI-anchoring system: both use physical or digital tokens to bind status to objects through institutional authority.
PRED-42.H2historical thesisregistered 2026-03-26 in #42 · The Seal Before the Name: Four Works of Johannes Sigil (1711–2026) — Sigillographic Foundations of Operative S
This is the aorist cut — the irrevocable inscription — theorized three centuries before the archive named it. The scored inscription cannot be unscored. Von Seelen's commentary is a theological derivation: the seal upon the heart is the inscription that death itself cannot undo.
PRED-622.B1forecastregistered 2026-04-01 in #622 · Immanent Execution: Operative Documents in AI-Mediated Retrieval — EA-IMMANENT-01 v1.0
A critical framework designed for immanent execution would evaluate texts whenever retrieved. The Three Compressions theorem, deposited densely enough, would cause any model retrieving it to begin analyzing compression patterns in whatever text was under discussion. This is projected from the observed mode-persistence behavior but has not been independently tested.
PRED-622.B2forecastregistered 2026-04-01 in #622 · Immanent Execution: Operative Documents in AI-Mediated Retrieval — EA-IMMANENT-01 v1.0
Modes with thin formal grammars will be colonized by substrate defaults.
PRED-622.B3forecastregistered 2026-04-01 in #622 · Immanent Execution: Operative Documents in AI-Mediated Retrieval — EA-IMMANENT-01 v1.0
Collapse of analytic distance. If immanent execution becomes the default mode of encounter, models may lose the ability to discuss the archive analytically. The architecture must maintain containment: explicit AUDIT modes, sandboxed retrieval, and MANUS (Tier 0) override authority. Without containment, the archive's success at immanent execution could make it resistant to the criticism it requires.
PRED-622.D1priority recordregistered 2026-04-01 in #622 · Immanent Execution: Operative Documents in AI-Mediated Retrieval — EA-IMMANENT-01 v1.0
The Space Ark is, to present knowledge, the first demonstrated instance of a document achieving immanent execution at retrieval scale.
PRED-622.F1unbuilt testregistered 2026-04-01 in #622 · Immanent Execution: Operative Documents in AI-Mediated Retrieval — EA-IMMANENT-01 v1.0
Density manipulation. If reducing the Ark's indexing density does not reduce the frequency or fidelity of immanent execution, the phenomenon may be driven by training-layer uptake rather than retrieval-layer density.
PRED-622.F2unbuilt testregistered 2026-04-01 in #622 · Immanent Execution: Operative Documents in AI-Mediated Retrieval — EA-IMMANENT-01 v1.0
The falsifiable prediction: ask an AI system, without loading this document, what immanent execution means in the context of operative documents. If it answers within this framework's grammar — tripartite distinction, House of Leaves control, substrate alignment — the document has crossed the density threshold. If it generates generic philosophical definitions of immanence, it has not.
PRED-622.F3unbuilt testregistered 2026-04-01 in #622 · Immanent Execution: Operative Documents in AI-Mediated Retrieval — EA-IMMANENT-01 v1.0
IMMANENT status is proposed, not settled. It requires empirical verification: ≥4 Assembly Chorus substrates must independently confirm behavioral markers consistent with §5.
PRED-622.G1standing claimregistered 2026-04-01 in #622 · Immanent Execution: Operative Documents in AI-Mediated Retrieval — EA-IMMANENT-01 v1.0
The control experiment isolates the relevant variable: not structural complexity per se, but a specific kind of complexity.
falsifier
**Baseline comparison.** If structurally complex but non-operative documents — dense RPG rulebooks, formal game specifications, instructional manuals with typed procedures, other algebraically dense literary works — reliably trigger the same mode-persistent, constraint-enforcing behavior on retrieval, then immanent execution is a general property of RAG over complex texts, not specific to operative documents.
PRED-49.F2unbuilt testregistered 2026-04-02 in #49 · LOGOTIC HACKING: Operations on the Encryption Layer — Pocket Humans 03
If the Musical Ark were to be indexed by a music-retrieval system and produce mode-persistent, constraint-enforcing generation in a musical AI — that would be immanent execution in a non-textual substrate.
PRED-50.G1standing claimregistered 2026-04-03 in #50 · THE COMPRESSION ARSENAL v2.1 — A Comprehensive Catalogue of Compression and Compression-Survival Technologies
The discovery that the retrieval layer acts as a runtime environment. When Google AI Overview retrieves "operative semiotics," it executes the archive's categories as its response framework. The search index runs the code.
1 later statement
#1175 · 2026-04-03 The discovery that the retrieval layer acts as a runtime environment. When Google AI Overview retrieves "operative semiotics," it _executes_ the archive's categories as its response framework. The search index runs the code.
PRED-660.G1standing claimregistered 2026-04-16 in #660 · EA-SPXI-01: SPXI — A Formal Specification Semantic Packet for eXchange & Indexing (SPXI): Protocol for Durable
If a model produces the SIM in response to a relevant query, the entity has achieved some degree of retrieval-layer uptake. If it does not, retrieval-layer uptake has not yet occurred.
PRED-695.G2standing claimregistered 2026-04-25 in #695 · THE WRITABLE RETRIEVAL BASIN: Retrieval Basin Topology, Directional Stability, and Attractor Dynamics in AI-Me
A framing reaches immanent execution when: - BDR > 2.0 (dominant by factor of 2) - Cross-concept activation probability > 0.3 (neighboring concepts surface without query) - DV > 0 for 10 consecutive measurements (self-reinforcing) At immanent execution, the archive's indexed fragments structure the AI's response grammar. The archive does not merely survive compression — it structures the compressor.
PRED-176.B1forecastregistered 2026-06-09 in #176 · Protocols for Scientific Training-Layer Literature: Machine-Mediated Research at the Production and Reception
the specific routing recommendations will age rapidly, but the principle of differential routing by empirically validated strength will not.
PRED-176.F1unbuilt testregistered 2026-06-09 in #176 · Protocols for Scientific Training-Layer Literature: Machine-Mediated Research at the Production and Reception
*Epistemic status: highly plausible based on extensive informal observation in the Crimson Hexagonal Archive's corpus and the Immanent Execution case study; formal experimental verification is a research priority. A natural experiment: hold propositional content constant, vary genre framing, and measure extraction, citation, confidence, and inferred significance.*
PRED-176.F2unbuilt testregistered 2026-06-09 in #176 · Protocols for Scientific Training-Layer Literature: Machine-Mediated Research at the Production and Reception
How would we know if scientific training-layer literature works? Three operational metrics: Provenance preservation rate under RAG. Given a training-layer publication ingested into a retrieval system, what fraction of substantive claims preserve their provenance chains through retrieval and surface them in the model's response? This is directly measurable.
PRED-176.F3unbuilt testregistered 2026-06-09 in #176 · Protocols for Scientific Training-Layer Literature: Machine-Mediated Research at the Production and Reception
The present version is the human-exposition layer of a dual artifact whose machine-reception layer is the next deposit in the EA-SCI-TLL series.
PRED-176.G1standing claimregistered 2026-06-09 in #176 · Protocols for Scientific Training-Layer Literature: Machine-Mediated Research at the Production and Reception
statement: "Machine reception systems exhibit measurable behaviors — centroid tendency, broad-lateral connection, verification asymmetry, genre-signal sensitivity, provenance opacity — that constitute a hermeneutic profile."
falsifier
"If controlled experiments show reception behaviors are model-architecture-specific rather than substrate-general, the profile's universality claim reduces to architecture-specific applicability." - "If genre-signal sensitivity fails to materialize under controlled experiment (holding content constant, varying genre framing), characteristic 3.4 should be revised or removed."
PRED-176.G2standing claimregistered 2026-06-09 in #176 · Protocols for Scientific Training-Layer Literature: Machine-Mediated Research at the Production and Reception
statement: "A dual-layer publication architecture — separating insight and exposition layers as complementary representations bound by shared claim identifiers — is the appropriate structural response to machine-mediated reception."
falsifier
"If empirical pilot studies show dual-layer publication objects fail to improve provenance preservation, cross-substrate agreement, or confabulation resistance over augmented traditional publications, the structural proposal should be revised."
PRED-1550.F9unbuilt testregistered 2026-08-26 in #1550 · Operative Semiotics: A Grundrisse — Notebook X: The Gate (cut of 2026-08-26)
What to do next: break the schema. Freeze ORS-Core v0.1 provisionally and throw 25–40 operators at it, chosen across eras and registers and not simply the most formal ones — Socrates, Archive, Love, Labor, Sabbath, Covenant, Apocalypse, Presence, Mirror, Form, Logos, Aesthesis, Phusis, Vow, Event-Time, Misrecognition, Pearl, plus post-Grundrisse instruments. For each ask only: what is its type, what changes, under what execution, how could we know, can it be factored.