ExperimentEXP-2026-0013v0.1

PACC-Lab v0.6 — hierarchical composed coordinate

A preregistered 95-dimensional composed coordinate — the validated task-state map's held-out decision quotient combined with the current relation pattern and a bounded set of interaction terms — predicts the Bayesian common-cause coordinate for all four families in both geometries: composed latent JS 0.0034–0.0073 versus 0.040–0.062 for shuffled/constant and 0.040–0.060 for a deliberately broken composition; counterfactual JS 0.0018–0.0047; four secondary seeds give direct 0.0 and composed 1.0 pass rates.

Research status: STABLE the current conclusions are relatively stable
Evidence level: E3 Repeated experiment
Result: POSITIVE
Data basis: SYNTHETIC Synthetic data and theoretical reasoning. Many now treat synthetic data as if it were real; this laboratory says the opposite deliberately — until a real hybrid model exists, an inference is only an inference, and theoretically possible is not actually possible.
Version: 0.1
Updated: 2026-09-09
Created: 2026-09-09
Domain: Model Representation, Evaluation, Formal AI
Program: PRG-2026-0001 Adaptive Epistemic Systems
Authors: Neo.K (EveMissLab)
AI collaborators: Sol (GPT-5.6, OpenAI ChatGPT)

Hypothesis

hypothesis: The latent probability coordinate is compositional relative to the non-probabilistic state: decision quotient + relation pattern, not the raw state, maps to the common-cause posterior.

Setup

model_ids: MOD-2026-0001 N0 — signed support
MOD-2026-0002 N1 — ordinal tournament
MOD-2026-0003 N2 — signed graph
MOD-2026-0004 N3 — constraint competition
MOD-2026-0005 Bayesian reference (exact posterior / Beta-Bernoulli / joint common-cause / HMM)
dataset_ids: DAT-2026-0001 PACC synthetic evidence worlds
benchmark_ids: BEN-2026-0001 PACC micro-lab protocol v0.1 (frozen gates)
software_environment: Python; deterministic seeded generators; no network, no LLM.

Runs

run_count

random_seeds

20260909
4 secondary seeds

controls

shuffled-target mapping
constant prediction
broken composition (v0.6+)
target-local refit (v0.7+)

metrics

verdict

HIERARCHICAL_COMPOSED_COORDINATE_RESCUES_LATENT_DEPENDENCE

direct_robust

0/4

composed_robust

4/4

moderate_N0

composed_latent_js: 0.003603
best_control: 0.061508
broken_composition: 0.058196
counterfactual: 0.00278

high_N0

composed_latent_js: 0.003335
best_control: 0.03973
broken_composition: 0.041238

Interpretation

interpretation: Destroying the train alignment between task quotient and relation pattern destroys most of the signal, so the rescue is not a snapshot fit. Reinterprets v0.5: the information was present; adding raw coordinates did not make the latent state a direct coordinate.

Limitations

limitations: Does not show all non-probabilistic states admit such a composition, that the coordinate transfers between geometries without refitting, or that clusters can be discovered unsupervised.

Reproduction

reproduction_instructions: Extract the version's FINAL bundle; python -m pytest -q; run the version's primary script with the recorded seed; docs/PACC_LAB_v0.N_RESULTS.md and docs/EXPERIMENT_PROTOCOL_v0.N.md are inside the bundle.

Results

Source	Relation	Target	Status	ID
`EXP-2026-0013` PACC-Lab v0.6 — hierarchical composed coordinate	`produces`	`RST-2026-0009` v0.6: composed coordinate 4/4, direct 0/4	`ACTIVE`	`REL-2026-0228`

Recorded fields

completed_at: 2026-09-09

Relations

Source	Relation	Target	Status	ID
`EXP-2026-0013` PACC-Lab v0.6 — hierarchical composed coordinate	`runs_on`	`SYS-2026-0002` PACC-Lab — micro-lab harness for the convergence conjecture	`ACTIVE`	`REL-2026-0217`
`EXP-2026-0013` PACC-Lab v0.6 — hierarchical composed coordinate	`uses_benchmark`	`BEN-2026-0001` PACC micro-lab protocol v0.1 (frozen gates)	`ACTIVE`	`REL-2026-0218`
`EXP-2026-0013` PACC-Lab v0.6 — hierarchical composed coordinate	`uses_dataset`	`DAT-2026-0001` PACC synthetic evidence worlds	`ACTIVE`	`REL-2026-0219`
`EXP-2026-0013` PACC-Lab v0.6 — hierarchical composed coordinate	`uses_model`	`MOD-2026-0001` N0 — signed support	`ACTIVE`	`REL-2026-0220`
`EXP-2026-0013` PACC-Lab v0.6 — hierarchical composed coordinate	`uses_model`	`MOD-2026-0002` N1 — ordinal tournament	`ACTIVE`	`REL-2026-0221`
`EXP-2026-0013` PACC-Lab v0.6 — hierarchical composed coordinate	`uses_model`	`MOD-2026-0003` N2 — signed graph	`ACTIVE`	`REL-2026-0222`
`EXP-2026-0013` PACC-Lab v0.6 — hierarchical composed coordinate	`uses_model`	`MOD-2026-0004` N3 — constraint competition	`ACTIVE`	`REL-2026-0223`
`EXP-2026-0013` PACC-Lab v0.6 — hierarchical composed coordinate	`uses_model`	`MOD-2026-0005` Bayesian reference (exact posterior / Beta-Bernoulli / joint common-cause / HMM)	`ACTIVE`	`REL-2026-0224`
`EXP-2026-0013` PACC-Lab v0.6 — hierarchical composed coordinate	`tests`	`THY-2026-0005` PACC conjecture — the four-level convergence ladder	`ACTIVE`	`REL-2026-0225`
`EXP-2026-0013` PACC-Lab v0.6 — hierarchical composed coordinate	`extends`	`EXP-2026-0012` PACC-Lab v0.5 — does a richer relation state rescue the latent coordinate?	`ACTIVE`	`REL-2026-0226`
`EXP-2026-0013` PACC-Lab v0.6 — hierarchical composed coordinate	`produced`	`ART-2026-0026` PACC-Lab v0.6 Composed Coordinate FINAL `artifact://evemisslab/adaptive-epistemic-systems/PACC-Lab_v0.6_Composed_Coordinate_FINAL.zip`	`ACTIVE`	`REL-2026-0227`
`EXP-2026-0013` PACC-Lab v0.6 — hierarchical composed coordinate	`produces`	`RST-2026-0009` v0.6: composed coordinate 4/4, direct 0/4	`ACTIVE`	`REL-2026-0228`
`EXP-2026-0014` PACC-Lab v0.7 — cross-geometry coordinate transfer without refitting	`extends`	`EXP-2026-0013` PACC-Lab v0.6 — hierarchical composed coordinate	`ACTIVE`	`REL-2026-0239`

History and provenance

Canonical URL

https://evemisslab.com/ai/experiments/EXP-2026-0013/

Machine-readable

/ai/experiments/EXP-2026-0013/index.json

Snapshot

AI-SNAPSHOT-v0.1-fe85b9694a45

Provenance

source: EveMissLab research collection: Adaptive Epistemic Systems (真本體論13)
extracted_by: Splice (Claude Code), reading the canonical UTF-8 sources and each lab's own result reports
extracted_at: 2026-09-11
generator: tools/extract_aes/extract.py
claim_boundary: status, evidence level and result type follow the source artifact's own stated claim boundary; nothing is upgraded beyond what the report supports