diff --git a/REPOSITORY_READINESS.md b/REPOSITORY_READINESS.md index abd550e..ddb6ebf 100644 --- a/REPOSITORY_READINESS.md +++ b/REPOSITORY_READINESS.md @@ -53,11 +53,11 @@ This checklist separates repository operations from architectural and ownership ## Current critical path -1. Review and preserve the E001 Stage A result as scoped E4 evidence. -2. Freeze a Stage B specification for noncausal pneumatic observables. -3. Retain Stage A configurations as controls during Stage B. -4. Test the zero-vacancy circulation boundary in a separate experiment or E001 v0.2. -5. Keep any causal exchange or escrow mechanism outside E001 v0.1. +1. Implement the E002 offline trace replayer without simulator imports. +2. Verify exact state reconstruction and byte-identical telemetry repetition. +3. Add optional shadow mode and prove Stage A trace noninterference. +4. Replay all 72 E001 runs and evaluate the frozen diagnostic thresholds. +5. Classify each pneumatic projection as retain, narrow, revise, or reject. ## E001 Stage A completion @@ -67,3 +67,13 @@ This checklist separates repository operations from architectural and ownership - [x] Raw evidence, summaries, provenance, and checksum preserved - [x] Directed results and emergent observations separated - [x] Exit decision recorded: proceed with noncausal Stage B instrumentation + +## E002 Stage B entrance criteria + +- [x] E002 has a standalone experiment specification +- [x] Source E001 evidence and archive digest are frozen +- [x] Offline replay and exact rational representation are recorded as an ADR +- [x] Sampling order and no-lookahead rule are explicit +- [x] Pneumatic variables are prohibited from changing Stage A transitions +- [x] Fixed diagnostic baselines and thresholds are declared before implementation +- [x] Retain, narrow, revise, and reject outcomes are defined diff --git a/docs/decisions/ADR__0002__OFFLINE_REPLAY_AND_EXACT_RATIONAL_TELEMETRY_FOR_E002__2026-07-14.md b/docs/decisions/ADR__0002__OFFLINE_REPLAY_AND_EXACT_RATIONAL_TELEMETRY_FOR_E002__2026-07-14.md new file mode 100644 index 0000000..9e3bc62 --- /dev/null +++ b/docs/decisions/ADR__0002__OFFLINE_REPLAY_AND_EXACT_RATIONAL_TELEMETRY_FOR_E002__2026-07-14.md @@ -0,0 +1,100 @@ +# ADR 0002 — Offline replay and exact rational telemetry for E002 + +**Date:** 2026-07-14 +**Status:** Trial +**Decision owner:** Travis Levi Streets +**Affected concepts or experiments:** `SW.EXPERIMENT.E002`, `SW.CONCEPT.FIELD_THERMODYNAMICS`, `SW.BRIDGE.CBF_PNEUMATIC` + +## Context + +E001 completed the discrete three-ring experiment and authorized noncausal Stage B instrumentation. E002 must determine whether pneumatic variables are coherent and useful without allowing them to change Stage A behavior or inherit unsupported physical meaning from the recovered Biological Pneumatic Engine. + +The decision is consequential because instrumentation placed inside the scheduler can accidentally become causal, floating-point edge cases can make canonical telemetry platform-sensitive, and a weighted energy score can hide arbitrary assumptions behind one persuasive number. + +## Decision drivers + +- Demonstrable noninterference with frozen E001 evidence +- Exact reconstruction and deterministic telemetry +- Explicit separation of capacity, fault, semantic, and authority states +- No wall-clock or platform-dependent modeled behavior +- Inspectable implementation using the existing Stage A environment +- Easy rejection or narrowing when a projection is redundant +- No premature calibration, learning, or causal feedback + +## Options considered + +### Instrument directly inside the Stage A simulator + +This would provide convenient access to internal state and enable shadow comparison. It also creates the greatest risk that observation changes ordering, state, or future design assumptions. It is deferred until an offline implementation defines canonical parity. + +### Offline replay with exact rational telemetry + +An offline replayer consumes immutable E001 JSON Lines traces, reconstructs state, and emits a separate telemetry stream. Integers and reduced rational numbers avoid floating-point ambiguity. Shadow execution can later compare the same derived records without becoming authoritative. + +### External numerical or fluid-simulation stack + +A numerical library could support continuous integration, differential equations, and richer visualization. E002 does not yet have evidence for continuous dynamics, calibrated units, or step-size selection. Adding that stack would enlarge the model before the observables have shown value. + +### Reuse the recovered Biological Pneumatic Engine + +The recovered engine contains pressure, viscosity, respiratory phase, and rhythm mechanisms. Its external and fallback paths are not behaviorally equivalent, its delays use wall time, and several mechanisms are diagnostic or unwired. Reuse would import precisely the ambiguities E002 is intended to isolate. + +## Decision + +E002 will use an **offline, read-only replay** of the frozen E001 evidence as its canonical implementation path. + +Canonical telemetry will: + +- use Python 3.13 and the standard library under the bounded ADR 0001 runtime trial; +- use integers and reduced rational objects, never canonical floating point; +- emit pre-tick and post-tick samples with an enforced no-lookahead rule; +- preserve typed blocked and unobserved states instead of infinity, `NaN`, or invented zeros; +- store telemetry separately from source traces; +- use fixed permeability 1 and no fitted coefficients; +- report space-time, dissipation proxy, coordination, and obligation components separately; +- represent phase as an integer event-count index with declared sensitivity periods; and +- compare fixed projections with simple raw scalar baselines. + +An optional shadow mode may be added only after offline replay is valid. Shadow mode must prove byte-identical Stage A traces and summaries with instrumentation enabled and disabled. + +No result from this trial authorizes a causal flow law. Stage C requires a new ADR and experiment specification. + +## Consequences + +### Expected benefits + +- The canonical E001 evidence remains immutable. +- Every derived value has an exact representation and source lineage. +- Instrumentation defects cannot silently change scheduler behavior. +- Redundant or misleading pneumatic projections can be rejected cheaply. +- The implementation remains independent of the recovered engine and third-party numerical stacks. + +### Costs and risks + +- Trace replay must reconstruct state that an in-process observer would access directly. +- Exact rational telemetry is more verbose than floating-point output. +- The first candidate flow law is intentionally simple and may perform poorly. +- Event-count phase is only a diagnostic index, not a continuous oscillator. +- Derived telemetry cannot add information beyond the complete source trace; usefulness is limited to projection, compression, and interpretation. + +## Validation and reversal + +The trial succeeds if all 72 source runs replay deterministically, every reconstruction and noninterference invariant passes, and the experiment can classify each projection as useful, redundant, or misleading without post hoc tuning. + +Revisit or reverse this decision if: + +- E001 state cannot be reconstructed from canonical traces without hidden simulator access; +- exact rational output makes the ordinary experiment impractical; +- shadow parity cannot be demonstrated; +- the fixed projections require calibrated continuous dynamics to be meaningfully evaluated; or +- a separate implementation reproduces the telemetry more clearly with another representation. + +## Independence and provenance + +The CBF–Pneumatic Bridge supplies the candidate vocabulary and sequencing. The Biological Pneumatic Engine supplies historical examples and negative engineering evidence. Neither artifact grants authority to alter E001, import a runtime, claim physical units, or treat phase as biological causality. + +## Links + +- [E002 experiment specification](../../experiments/E002/EXPERIMENT__E002__NONCAUSAL_PNEUMATIC_INSTRUMENTATION__v0.1__2026-07-14.md) +- [E001 Stage A report](../../experiments/E001/results/reports/E001__STAGE_A_RESULTS__v0.1__2026-07-14.md) +- [CBF–Pneumatic Computation Bridge](../bridges/COMPUTATIONAL_ARCHITECTURE__BRIDGE__CBF_AND_PNEUMATIC_COMPUTATION__v0.1__2026-07-13.md) diff --git a/docs/decisions/README.md b/docs/decisions/README.md index 99499c3..f7b5e15 100644 --- a/docs/decisions/README.md +++ b/docs/decisions/README.md @@ -31,3 +31,4 @@ Copy [the ADR template](ADR_TEMPLATE.md), fill every section, and link supersedi ## Current decisions - [ADR 0001 — Python standard library for E001](ADR__0001__PYTHON_STANDARD_LIBRARY_FOR_E001__2026-07-13.md) — **Trial** +- [ADR 0002 — Offline replay and exact rational telemetry for E002](ADR__0002__OFFLINE_REPLAY_AND_EXACT_RATIONAL_TELEMETRY_FOR_E002__2026-07-14.md) — **Trial** diff --git a/experiments/E002/EXPERIMENT__E002__NONCAUSAL_PNEUMATIC_INSTRUMENTATION__v0.1__2026-07-14.md b/experiments/E002/EXPERIMENT__E002__NONCAUSAL_PNEUMATIC_INSTRUMENTATION__v0.1__2026-07-14.md new file mode 100644 index 0000000..4bd62c5 --- /dev/null +++ b/experiments/E002/EXPERIMENT__E002__NONCAUSAL_PNEUMATIC_INSTRUMENTATION__v0.1__2026-07-14.md @@ -0,0 +1,406 @@ +# E002 — Noncausal Pneumatic Instrumentation + +**Artifact class:** Observational experiment specification +**Version:** v0.1 +**Date:** 2026-07-14 +**Status:** Specified — E1; implementation not yet evaluated +**Registry identifier:** `SW.EXPERIMENT.E002` +**Stage:** B — instrumented pneumatic variables +**Decision dependency:** [ADR 0002](../../docs/decisions/ADR__0002__OFFLINE_REPLAY_AND_EXACT_RATIONAL_TELEMETRY_FOR_E002__2026-07-14.md) +**Evidence dependency:** [E001 Stage A report](../E001/results/reports/E001__STAGE_A_RESULTS__v0.1__2026-07-14.md) + +## 1. Purpose + +E002 asks whether the pneumatic vocabulary proposed by the CBF–Pneumatic Bridge can be made coherent, deterministic, and diagnostically useful when applied to the completed E001 traces. + +It introduces no new scheduler and changes no transition. Pressure, vacancy, resistance, conductance, candidate flow, phase, occupancy-time, and dissipation proxy are computed from immutable Stage A evidence by an offline replayer. A later shadow mode may compute the same telemetry during an E001 rerun, but the Stage A canonical trace must remain byte-identical. + +E002 is a new experiment rather than an amendment to E001. E001 v0.1 and its unfavorable full-ring result remain frozen. + +## 2. Primary question + +Can a typed pneumatic observability layer describe Stage A field behavior without changing it, collapsing semantic distinctions, or adding a misleading scalar metaphor? + +## 3. Constitutional boundary + +E002 telemetry is read-only. No derived value may: + +- authorize or reject a proposal; +- create, consume, renew, or release capacity credit; +- open or close a semantic aperture; +- change queue order, service order, transfer order, or terminal state; +- override identity, schema, provenance, leases, isolation, or prohibited states; +- add an atomic cyclic exchange or unrestricted bypass; or +- become input to the Stage A simulator during the canonical experiment. + +Any implementation path that violates this boundary belongs to a new causal experiment and cannot be reported as E002. + +## 4. Hypotheses + +### E002-H1 — Exact noninterference + +Offline instrumentation and shadow instrumentation leave every canonical E001 transition and summary unchanged. + +**Falsified if:** enabling telemetry changes a Stage A event, event order, trace digest, summary digest, terminal count, or invariant result. + +### E002-H2 — Coherent reconstruction + +The replayer reconstructs loop occupancy, work location, readiness, obligations, loop mode, and interlock state exactly enough to reproduce every declared Stage A accounting identity. + +**Falsified if:** reconstructed state disagrees with a trace transition, produces an impossible rational value, loses a work identifier, or requires unrecorded information to explain a promoted transition. + +### E002-H3 — Typed diagnostic separation + +Capacity backpressure, fault isolation, semantic rejection, stale authority, and reciprocal debt remain distinguishable telemetry channels rather than being collapsed into one pressure scalar. + +**Falsified if:** an absent measurement is serialized as zero, an invalid proposal appears transport-eligible because of high pressure, or two constitutionally different causes become indistinguishable in the canonical telemetry. + +### E002-H4 — Diagnostic usefulness + +At least one fixed pneumatic projection provides a stable, scenario-general diagnostic of a declared Stage A outcome while remaining explicit about its raw queue inputs. + +**Supported if:** a fixed pre-tick pneumatic score reaches ROC-AUC at least 0.70 for a declared next-step target in at least six of eight scenarios and exceeds the receiver-occupancy scalar baseline by at least 0.05 in the same pooled comparison. + +**Not supported if:** the projections are coherent but fail the threshold, merely rename raw occupancy, or reverse meaning across scenarios. This is a valid negative result and blocks any claim that Stage B has demonstrated constitutive value. + +## 5. Source evidence and replay matrix + +The primary dataset is the complete E001 Stage A evidence set: + +| Property | Frozen value | +| --- | --- | +| Source experiment | `SW.EXPERIMENT.E001` | +| Source commit | `81e7f859f71425bdba7603a61566f9fb47c116f9` | +| Canonical archive SHA-256 | `584cc15bfeb3cc74dec9d9069cde26e1abaa6f1350c0aa12ae10a9784bd1663b` | +| Configurations | central, global barrier, local CBF | +| Scenarios | 8 frozen E001 scenarios | +| Seeds | 17, 29, 43 | +| Core replays | 72 | + +No source run may be omitted. E002 produces one telemetry stream and one analysis summary for each source run. Phase sensitivity is derived from the same stream for periods 2, 3, 4, and 6; it does not multiply or mutate source runs. + +The noncanonical one-vacancy perturbation recorded in E001 Observation O001 is not part of E002's canonical matrix. + +The canonical replayer must consume the preserved archive and must not import `superloop_e001.simulator`, `superloop_e001.workloads`, or inspect live simulator objects. Shared canonical-JSON and digest helpers may be reimplemented or placed in a representation-only module that has no transition authority. + +## 6. Replay state + +For loop `i` at the start of tick `t`, reconstruct: + +```text +Q_i(t) = { + occupancy, + capacity, + active_work_ids, + processed_work_ids, + ready_transfer_count, + oldest_wait_age, + outstanding_obligations, + process_count, + mode +} +``` + +For directed interlock `i -> j`, reconstruct: + +```text +I_ij(t) = { + source_ready, + receiver_vacancy, + receiver_mode, + proposal_eligibility, + lease_state, + current_wait_age, + committed_flow +} +``` + +`proposal_eligibility` includes only facts already known before the tick: active work, completed local processing, valid identity and schema, provenance, freshness of any pre-existing lease, and non-isolated participants. It means that work may seek a transfer; it is not a validated aperture or permission to commit. Receiver vacancy is represented separately as capacity conductance. + +## 7. Sampling order and no-lookahead rule + +Each source tick produces two samples: + +1. **Pre-tick sample.** Reconstructed exclusively from events before tick `t`. It may be used to score events at tick `t` or later. +2. **Post-tick sample.** Reconstructed after every canonical event at tick `t` has been applied in event order. It describes outcome state and may not be used as a predictor for the same tick. + +Targets are taken from canonical Stage A events. No pre-tick observable may read a same-tick event, terminal outcome, or future state. Lookahead makes the run invalid even when aggregate metrics appear favorable. + +## 8. Exact numeric representation + +Canonical telemetry uses integers and reduced rational numbers: + +```json +{"numerator": 3, "denominator": 8} +``` + +Rules: + +- denominator is strictly positive; +- zero is encoded as `0/1`; +- numerator and denominator share no common factor except 1; +- floating-point values, `NaN`, and infinity are prohibited from canonical telemetry; and +- unavailable measurements use a typed state such as `not_observed`, never numeric zero. + +Floating-point renderings may be generated for exploratory charts but are excluded from canonical equality and digests. + +## 9. Declared observables + +### 9.1 Loop occupancy and vacancy + +For loop capacity `C_i` and occupancy `q_i(t)`: + +```text +congestion_i(t) = q_i(t) / C_i +vacancy_i(t) = (C_i - q_i(t)) / C_i +``` + +Both lie in `[0, 1]` and sum exactly to 1. + +### 9.2 Typed pressure vector + +E002 retains separate channels: + +```text +pressure_i(t) = { + congestion: q_i / C_i, + transfer_demand: min(ready_transfer_count_i, C_i) / C_i, + wait_strain: min(oldest_wait_age_i, deadline) / deadline, + verification_debt: min(outstanding_obligations_i, C_i) / C_i, + uncertainty: not_observed, + trust: not_observed +} +``` + +`uncertainty` and `trust` are explicitly unobserved because E001 did not model them. Authorization is never a pressure component. + +### 9.3 Capacity resistance and conductance + +Receiver capacity conductance is: + +```text +K_ij(t) = vacancy_j(t) +``` + +When receiver vacancy is positive, the corresponding dimensionless resistance is: + +```text +R_ij(t) = 1 / K_ij(t) = C_j / (C_j - q_j(t)) +``` + +When vacancy is zero, resistance is encoded as: + +```json +{"state": "blocked", "reason": "zero_receiver_vacancy"} +``` + +It is not encoded as infinity. Fault isolation and semantic ineligibility are separate states and must not be mislabeled as capacity resistance. + +### 9.4 Congestion potential and gradient + +The primary scalar transport potential is deliberately minimal: + +```text +U_i(t) = congestion_i(t) +G_ij(t) = max(U_i(t) - U_j(t), 0) +``` + +This projection may prove uninformative when equal occupancies exchange work. Such a result is evidence against the projection, not permission to tune it after inspection. + +### 9.5 Candidate flow + +With fixed permeability `k_ij = 1` and time step `dt = 1`: + +```text +F*_ij(t) = proposal_eligibility_ij(t) * K_ij(t) * G_ij(t) +``` + +`F*` is a counterfactual diagnostic, not a reconstructed validation decision. It does not cap or schedule committed Stage A flow. The canonical committed flow is the count of `transfer_commit` events on the interlock during the tick. + +### 9.6 Occupancy-time and field vacancy + +```text +space_time = sum_t sum_i q_i(t) +field_vacancy(t) = sum_i (C_i - q_i(t)) +``` + +These are exact discrete sums. They do not represent wall-clock cost or physical volume. + +### 9.7 Dissipation proxy + +E002 reports an unweighted dimensionless proxy: + +```text +D*_ij(t) = F*_ij(t) * G_ij(t) +``` + +Space-time, dissipation proxy, coordination events, and obligation-time remain separate components. E002 does not combine them into a weighted cost functional because no calibration evidence exists for coefficients. + +### 9.8 Event phase + +Let `N_i(t)` be the count of canonical `process` events at loop `i` before tick `t`. For declared period `P`: + +```text +phase_index_i(t, P) = N_i(t) mod P +phase_distance_ij(t, P) = min( + (phase_i - phase_j) mod P, + (phase_j - phase_i) mod P +) +``` + +The primary period is `P = 3`; sensitivity periods are 2, 4, and 6. Canonical telemetry stores integer phase indices and distances, not trigonometric floats. Event phase is an activity descriptor, not respiration, entrainment, or a clock. + +### 9.9 Zero-vacancy cycle diagnostic + +For the three-ring topology: + +```text +zero_vacancy_cycle(t) = 1 + iff field_vacancy(t) = 0 + and every directed interlock has ready transfer demand +``` + +This diagnostic observes the E001 O001 boundary. It cannot create cyclic credit or authorize simultaneous exchange. + +## 10. Fixed diagnostic comparisons + +All scores are computed from pre-tick samples. + +### Target T1 — Wait or credit denial at tick `t` + +- Raw scalar baseline: receiver congestion. +- Pneumatic score: `1 - F*_ij(t)` for an interlock with ready demand. + +### Target T2 — Transfer commit at tick `t` + +- Raw scalar baseline: receiver vacancy. +- Pneumatic score: `F*_ij(t)`. + +### Target T3 — Healthy-loop stall within five ticks + +- Raw scalar baseline: current healthy-loop congestion. +- Pneumatic score: mean incoming capacity resistance, with blocked capacity represented by score 1 and finite resistance normalized as `1 - K`. + +ROC-AUC uses average ranks for ties and is reported per scenario and pooled across seeds. Undefined AUC, including a target with only one class, is serialized as `null` and cannot support H4. + +The pneumatic scores are deterministic transformations of Stage A state and therefore do not add independent information to the full raw trace. H4 tests whether they provide a useful fixed projection relative to a single raw scalar, not whether they create information. + +## 11. Telemetry contract + +Each telemetry record contains exactly these top-level fields: + +```text +schema_version +source_run_id +source_trace_digest +sample_id +simulation_tick +sample_phase +configuration +scenario +seed +loop_id +interlock_id +source_event_ids +reconstruction_state +pressure_vector +congestion_potential +receiver_conductance +receiver_resistance +positive_gradient +candidate_flow +committed_flow +dissipation_proxy +field_vacancy +phase_indices +zero_vacancy_cycle +instrumentation_checks +``` + +Records serialize as canonical JSON Lines with sorted keys. Unused context is `null`; required keys are never omitted. Execution platform, wall time, and memory use belong in a separate noncanonical execution record. + +## 12. Hard instrumentation invariants + +1. Source trace and summary digests are unchanged. +2. Offline replay emits no Stage A transition and writes no source evidence file. +3. Reconstructed occupancy and work ledgers agree with every source event. +4. Every rational is reduced, bounded where declared, and has a positive denominator. +5. Missing uncertainty or trust is `not_observed`, never zero. +6. Capacity blockage, fault isolation, and semantic rejection remain distinct states. +7. A pre-tick sample contains no same-tick or future information. +8. Every telemetry record links to one source run and source trace digest. +9. Pressure, flow, resistance, dissipation, and phase never authorize a transition. +10. Repeated instrumentation produces byte-identical canonical telemetry and analysis summaries. + +Any violation invalidates the affected run and blocks interpretation. + +## 13. Analysis outputs + +Each source run produces: + +- canonical loop and interlock telemetry; +- reconstruction and noninterference checks; +- component ranges and null-state counts; +- T1 through T3 outcome counts and AUC values; +- occupancy-time and obligation-time components; +- candidate-flow and committed-flow correspondence; +- phase-distance distributions for periods 2, 3, 4, and 6; +- zero-vacancy cycle intervals; and +- a digest-linked summary. + +The aggregate report separates: + +1. hard validity and noninterference; +2. directed evidence for H1 through H4; +3. coherent-but-redundant measurements; +4. misleading or rejected projections; and +5. emergent observations. + +## 14. Outcome classes + +### Retain + +All hard invariants pass, H1 through H3 are supported, and H4 meets its predeclared diagnostic threshold. Retain the useful variables as noncausal observability concepts. + +### Narrow + +The layer is coherent and noninterfering, but H4 is not supported or phase/dissipation measures are redundant. Preserve the negative result and retain only the variables with clear diagnostic meaning. + +### Revise + +Reconstruction, numeric representation, sampling order, or target definition is defective. Correct the implementation or issue a new specification before rerunning. + +### Reject + +The pneumatic projections systematically misdescribe the field, collapse constitutional distinctions, or require post hoc tuning to appear useful. Reject or retire those projections without weakening the CBF contract. + +No E002 outcome authorizes Stage C automatically. A causal constitutive layer requires a new frozen experiment with equivalent-work controls and explicit safety review. + +## 15. Evidence locations + +- Configuration: `experiments/E002/config/` +- Raw telemetry: `experiments/E002/results/raw/` +- Run summaries: `experiments/E002/results/summary/` +- Aggregate reports: `experiments/E002/results/reports/` +- Emergent observations: `experiments/E002/observations/` + +## 16. Non-claims + +E002 does not claim: + +- physical, pneumatic, biological, or thermodynamic equivalence; +- that pressure is information, authority, correctness, or trust; +- causal scheduling or improved workload behavior; +- conservation of semantic information; +- useful entrainment or neural-style learning; +- calibrated energy, monetary cost, or hardware-resource prediction; +- generality beyond the E001 evidence set; or +- permission to repair the full-ring boundary through hidden coordination. + +## Internal references + +- [CBF–Pneumatic Computation Bridge](../../docs/bridges/COMPUTATIONAL_ARCHITECTURE__BRIDGE__CBF_AND_PNEUMATIC_COMPUTATION__v0.1__2026-07-13.md) +- [Biological Pneumatic Engine paper](../../docs/whitepapers/PNEUMATIC_COMPUTATION__WHITEPAPER__BIOLOGICAL_PNEUMATIC_ENGINE__v0.1__2026-07-10.md) +- [Minimal Interlock Contract](../../specs/CBF__SPEC__MINIMAL_INTERLOCK_CONTRACT__v0.1__2026-07-10.md) +- [E001 Observation O001](../E001/observations/E001__OBSERVATION__FULL_RING_CIRCULATION_ASYMMETRY__v0.1__2026-07-14.md) +- [Research Governance](../../GOVERNANCE.md) diff --git a/experiments/E002/config/INSTRUMENTATION_CONTRACT__E002__v0.1__2026-07-14.json b/experiments/E002/config/INSTRUMENTATION_CONTRACT__E002__v0.1__2026-07-14.json new file mode 100644 index 0000000..09be365 --- /dev/null +++ b/experiments/E002/config/INSTRUMENTATION_CONTRACT__E002__v0.1__2026-07-14.json @@ -0,0 +1,81 @@ +{ + "analysis": { + "auc": { + "minimum_pneumatic_auc": { + "denominator": 10, + "numerator": 7 + }, + "minimum_scenario_count": 6, + "minimum_scalar_baseline_lift": { + "denominator": 20, + "numerator": 1 + }, + "tie_policy": "average_rank" + }, + "targets": [ + "wait_or_credit_denial_same_tick", + "transfer_commit_same_tick", + "healthy_loop_stall_within_five_ticks" + ] + }, + "canonical_numeric_representation": { + "floating_point_allowed": false, + "infinity_allowed": false, + "rational_fields": [ + "numerator", + "denominator" + ], + "zero": { + "denominator": 1, + "numerator": 0 + } + }, + "experiment": "SW.EXPERIMENT.E002", + "instrumentation": { + "candidate_flow": "proposal_eligibility * receiver_conductance * positive_gradient", + "congestion_potential": "occupancy / capacity", + "dissipation_proxy": "candidate_flow * positive_gradient", + "permeability": { + "denominator": 1, + "numerator": 1 + }, + "phase_periods": [ + 2, + 3, + 4, + 6 + ], + "primary_phase_period": 3, + "receiver_conductance": "receiver_vacancy / receiver_capacity", + "receiver_resistance": "receiver_capacity / receiver_vacancy", + "sampling_phases": [ + "pre_tick", + "post_tick" + ], + "time_step": 1 + }, + "prohibited_effects": [ + "authorization", + "capacity_credit_mutation", + "queue_reordering", + "service_reordering", + "transfer_reordering", + "terminal_state_mutation", + "cyclic_exchange_authority", + "stage_a_trace_mutation" + ], + "schema_version": "e002.instrumentation-contract.v1", + "source_evidence": { + "archive_sha256": "584cc15bfeb3cc74dec9d9069cde26e1abaa6f1350c0aa12ae10a9784bd1663b", + "configurations": 3, + "experiment": "SW.EXPERIMENT.E001", + "runs": 72, + "scenarios": 8, + "seeds": [ + 17, + 29, + 43 + ], + "source_commit": "81e7f859f71425bdba7603a61566f9fb47c116f9" + } +} diff --git a/experiments/E002/observations/README.md b/experiments/E002/observations/README.md new file mode 100644 index 0000000..cf1f955 --- /dev/null +++ b/experiments/E002/observations/README.md @@ -0,0 +1,9 @@ +# E002 Emergent Observations + +This directory records behavior not predicted by E002's directed hypotheses. + +Each observation must identify the source E001 run, source trace digest, E002 implementation commit, telemetry interval, description before interpretation, ordinary explanations, and the smallest discriminating follow-up. + +Observations begin as Seed / E0. They cannot change E002 projections, thresholds, or outcome classes retroactively. + +No E002 observations have been recorded yet. diff --git a/experiments/E002/results/README.md b/experiments/E002/results/README.md new file mode 100644 index 0000000..ba75042 --- /dev/null +++ b/experiments/E002/results/README.md @@ -0,0 +1,15 @@ +# E002 Results + +This directory will preserve noncausal telemetry derived from the frozen E001 evidence set. + +## Layout + +- `raw/` — canonical JSON Lines telemetry, stored separately from E001 traces +- `summary/` — one machine-readable instrumentation summary per source run +- `reports/` — aggregate validity, diagnostic, and interpretation reports + +Every result set must identify the E002 specification and configuration digests, implementation commit, source E001 run and trace digest, invocation, runtime, and every invalid or omitted run. + +Generated telemetry may not overwrite E001 evidence. Corrections require a new result identifier and preserved supersession record. + +No E002 evidence has been produced yet. diff --git a/experiments/README.md b/experiments/README.md index 9428b6d..8e25254 100644 --- a/experiments/README.md +++ b/experiments/README.md @@ -47,3 +47,21 @@ The CBF configuration remains bounded and live under the declared normal and deg ### Result The local CBF configuration preserved all hard invariants and reduced fault radius relative to the global barrier under failed-neighbor S5. It incurred higher coordination cost and exposed an exactly-full-ring liveness boundary. The Stage A report recommends proceeding to noncausal pneumatic instrumentation while preserving that limitation. + +## E002 — Noncausal Pneumatic Instrumentation + +**Status:** [Specified — E1](E002/EXPERIMENT__E002__NONCAUSAL_PNEUMATIC_INSTRUMENTATION__v0.1__2026-07-14.md) + +**Registry ID:** `SW.EXPERIMENT.E002` + +### Question + +Can pressure, resistance, conductance, candidate flow, occupancy-time, dissipation, and event phase become coherent and useful observables of the frozen E001 field without changing any transition? + +### Boundary + +E002 replays E001 evidence offline and emits a separate exact-rational telemetry stream. The observables cannot authorize work, create credit, change order, repair circular wait, or become causal input. + +### Exit + +Classify the instrumentation as retain, narrow, revise, or reject. No outcome automatically authorizes a causal constitutive layer. diff --git a/registry/CONCEPT_REGISTRY.md b/registry/CONCEPT_REGISTRY.md index ce4781c..10f94fd 100644 --- a/registry/CONCEPT_REGISTRY.md +++ b/registry/CONCEPT_REGISTRY.md @@ -12,6 +12,7 @@ This registry is the identity and status index for foundational Superloop concep | `SW.PROTOTYPE.BPE` | Biological Pneumatic Engine | Prototyped / archived input | E2-limited | Recovered PDP v2 prototype; not a Superloop runtime dependency | | `SW.BRIDGE.CBF_PNEUMATIC` | CBF–Pneumatic Bridge | Formalized | E1 | Controlled synthesis: CBF constitution plus optional pneumatic dynamics | | `SW.EXPERIMENT.E001` | Three-Ring Bounded Flow | Completed | E4-scoped | Repeated 72-run Python 3.13 comparison; result is limited to the declared three-ring model | +| `SW.EXPERIMENT.E002` | Noncausal Pneumatic Instrumentation | Specified | E1 | Offline Stage B replay of E001 with exact read-only field telemetry | ## Naming rule