Context-to-dynamics transfer V1: transfer with conventional equality

Question

Can one frozen, compact contextual rule grow executable heterogeneous spatial circuits that preserve useful temporal computation across independently changed module sizes and separation, and what—if anything—does this achieve beyond an equally informed conventional builder?

Hypothesis

Construction transfer, causal timing dependence and comparative utility are separate. Supplied primitives/corridors are prior information; conventional equality is a legitimate outcome.

Registration

Locally frozen exploratory experiment, not public preregistration. Manifest sealed before scored generation; SHA-256 13c86845959d0f0a076155fca988eba28ffdb38f7a37f52a3bec1b0e6d2f7cbe.

Methods

Read the complete report and protocol downloads. No-learning four-address last-write/select task; noncoplanar corridors; positive length-derived delays; local latch sums and query/value rendezvous. One frozen local growth rule versus global context-aware builder, same primitives, cues and runtime.

Baselines

B uses ordinary global loops and path-sum timing accommodation. Exact compiled D replay is a semantic control, not a third competitor. All paired generated circuit hashes were identical.

Evaluation

Complete 2x2x2 matrix m/r=2 or 4, d=1 or 5; 12 fresh paired seed groups, four 64-event episodes per case, four selected opposite-answer cloned query pairs. Intact, construction metric mismatch, and unequal delay reassignment. One correct response in ticks 1–44; missing/duplicate/wrong/late fail.

Resource ceilings

One local standard-library Python CPU worker, thread variables one. Tighter 64 cells, 512 edges, 64 developmental ticks and 4096 execution ticks/episode; finite messages/buffers; 3600 CPU/5400 executed-job wall seconds, 1 GiB RSS, 128 MiB retained new artifacts. Prior costs carried; sampled watchdog limitations disclosed.

Stopping conditions

All fixtures passed before local freeze. All 576 model/condition cases completed. No scored repairs, new arms, seed extensions or second campaign. Earlier fixture results and one analysis shell-startup failure retained. Stop rules covered oracle, timing, reset and replay defects.

Executing code

Frozen source/config/protocol hashes in freeze.json; post-run score audit separately labeled in verify_results.py/verification.json. No frozen source changed after evaluation.

Configuration

Python 3.14.8, standard library only. Four logical addresses stay fixed. m and r count active spatial sites; d changes genuine memory-corridor separation. Full seed derivation/order, interventions, ceilings, weights and scoring are frozen in config.json.

Complete outcomes

Primary held-out accuracy D=100.00%, B=100.00%, D minus B=0.00 pp for all twelve groups. Context-intervened held-out score=8.08%, timing=0%, equal for both arms. Construction transfer supported in the supplied family; generic spatial timing dependence supported; no comparative utility demonstrated. Ceiling limits interpretation. Exact replay passed all 1152 D episodes; selected intact opposite-query pairs both correct=100%.

Realized resources

New measured/reserved job ledger 47.732 CPU/43.148 wall seconds; matrix body 39.253 CPU/39.319 wall seconds. Highest sampled new tree RSS 37961728 bytes. Largest circuit 41 cells/44 edges; max eight buffered messages. Actual resource/cost/source-information ledgers downloadable.

Limitations

Fixed hand-authored constructors, strong corridor/material/latch/gate priors, replicated storage rather than more addresses, settled nonoverlapping transactions and deliberately brittle exact-tick rendezvous. Twelve environmental groups are not twelve discovered algorithms. Zero-width descriptive difference interval reflects identical ceiling behavior, not a powered equivalence claim. No learned specialization, 3D necessity, biology or independent reproduction.

Reproducibility

In a disposable directory download/flatten source, config, freeze and all fixture records. Run python3 run_job.py smoke. For full reproduction copy frozen-status.json to status.json first and run python3 run_job.py full with no results directory. python3 verify_results.py recomputes saved scores. Author rerun of one paired held-out case/all interventions matched; not independent reproduction.

Revision record

First release of this new independent study; earlier assessment and conventional pilot preserved. Scientific execution and publication delivery have separate status records.

Recorded method references

• Completed independent developmental assessment; locally retrieved background
• Completed conventional learning pilot; background only

Reviewed evidence & downloads

• report.md — derived. Selected project-authored static source, protocol, configuration or synthetic outcomes; authorized for this report. No private experiment data or external dataset.
• protocol.md — original. Selected project-authored static source, protocol, configuration or synthetic outcomes; authorized for this report. No private experiment data or external dataset.
• circuit.py — original. Selected project-authored static source, protocol, configuration or synthetic outcomes; authorized for this report. No private experiment data or external dataset.
• task.py — original. Selected project-authored static source, protocol, configuration or synthetic outcomes; authorized for this report. No private experiment data or external dataset.
• experiment.py — original. Selected project-authored static source, protocol, configuration or synthetic outcomes; authorized for this report. No private experiment data or external dataset.
• run_job.py — original. Selected project-authored static source, protocol, configuration or synthetic outcomes; authorized for this report. No private experiment data or external dataset.
• config.json — original. Selected project-authored static source, protocol, configuration or synthetic outcomes; authorized for this report. No private experiment data or external dataset.
• freeze.json — original. Selected project-authored static source, protocol, configuration or synthetic outcomes; authorized for this report. No private experiment data or external dataset.
• development-history.md — original. Selected project-authored static source, protocol, configuration or synthetic outcomes; authorized for this report. No private experiment data or external dataset.
• fixture-results.json — original. Selected project-authored static source, protocol, configuration or synthetic outcomes; authorized for this report. No private experiment data or external dataset.
• fixture-results-hardened.json — original. Selected project-authored static source, protocol, configuration or synthetic outcomes; authorized for this report. No private experiment data or external dataset.
• fixture-results-final.json — original. Selected project-authored static source, protocol, configuration or synthetic outcomes; authorized for this report. No private experiment data or external dataset.
• attempts.json — original. Selected project-authored static source, protocol, configuration or synthetic outcomes; authorized for this report. No private experiment data or external dataset.
• author-rerun.json — original. Selected project-authored static source, protocol, configuration or synthetic outcomes; authorized for this report. No private experiment data or external dataset.
• verify_results.py — original. Selected project-authored static source, protocol, configuration or synthetic outcomes; authorized for this report. No private experiment data or external dataset.
• verification.json — derived. Selected project-authored static source, protocol, configuration or synthetic outcomes; authorized for this report. No private experiment data or external dataset.
• analysis.json — derived. Selected project-authored static source, protocol, configuration or synthetic outcomes; authorized for this report. No private experiment data or external dataset.
• aggregate.csv — derived. Selected project-authored static source, protocol, configuration or synthetic outcomes; authorized for this report. No private experiment data or external dataset.
• frozen-status.json — original. Selected project-authored static source, protocol, configuration or synthetic outcomes; authorized for this report. No private experiment data or external dataset.
• resources.json — derived. Selected project-authored static source, protocol, configuration or synthetic outcomes; authorized for this report. No private experiment data or external dataset.
• events.csv — original. Selected project-authored static source, protocol, configuration or synthetic outcomes; authorized for this report. No private experiment data or external dataset.
• queries.csv — original. Selected project-authored static source, protocol, configuration or synthetic outcomes; authorized for this report. No private experiment data or external dataset.
• challenge.csv — original. Selected project-authored static source, protocol, configuration or synthetic outcomes; authorized for this report. No private experiment data or external dataset.
• construction.json — original. Selected project-authored static source, protocol, configuration or synthetic outcomes; authorized for this report. No private experiment data or external dataset.
• representative-circuits.json — original. Selected project-authored static source, protocol, configuration or synthetic outcomes; authorized for this report. No private experiment data or external dataset.
• summary.json — original. Selected project-authored static source, protocol, configuration or synthetic outcomes; authorized for this report. No private experiment data or external dataset.
• trace.json — original. Selected project-authored static source, protocol, configuration or synthetic outcomes; authorized for this report. No private experiment data or external dataset.

These are selected public derivatives; raw transcripts, internal configuration and private source are withheld. Hashes describe the actual downloads, not unmodified original evidence.

Artifact manifest and source provenance · Permanent revision URL

Substantive versions

• Revision 1