# Sīdus E1 RC4 — Canonical Claim Freeze

**Freeze date:** 4 November 2026  
**Status:** FROZEN CLAIM RECORD  
**Scientific disposition:** PASS  
**Scope rule:** define → test locally → replicate → only then generalize

## Canon-aligned headline

**Sīdus E1 passed its frozen test at the declared scope: in the tested synthetic fixture, several internally generated monitoring channels placed under a genuinely shared rate constraint exhibited preregistered, predictable scarcity and allocation relations.**

This is a local E1 result. It is not yet a general law of cognition, a result about humans or brains, proof of strict metarepresentation, or an empirical demonstration of a metarepresentational ecology.

## Frozen scientific claim

Within the frozen Sīdus E1 `U2`, `eta=0`, Arm S, lead `k=4` synthetic fixture, 501 preregistered training seeds (`9201`–`9701`) were each paired across six fixed conditions: base `C=4.5`, base `C=0.8`, shift `C=4.5`, shift `C=0.8`, alpha `C=0.8`, and beta `C=0.8`. All 3,006 registered conditions completed.

Under the prospective composite analysis over all 501 seeds:

1. **T1 — binding scarcity:** reducing the available budget from `C=4.5` to `C=0.8` produced a normalized paired return loss greater than the registered `0.05` margin in both base and shift environments.
2. **T2 — value-sensitive displacement:** at fixed binding budget `C=0.8`, the registered value shift moved more than `0.05` of total rate share toward source 2; because shares sum to one, this mass was displaced from the other sources.
3. **U — causal total reliance:** replacing each monitored source with an independent draw from the trained `q(S_v | I)` produced a normalized return loss greater than the registered `0.02` margin in all six environment-by-source cells.
4. **V — allocation prediction:** the frozen reference-v2 allocation prediction had more than `0.05` lower L1 allocation error than every eligible registered comparator in the eleven alpha/beta comparisons, under the frozen Holm-controlled family.

T1, T2, U, and V were each **CONFIRMED**. Under the frozen rule—PASS iff all four are confirmed; FAIL iff any is disconfirmed; otherwise HOLD—the overall result is **PASS**.

## What this establishes

- The shared constraint was empirically consequential in both registered training environments.
- Allocation responded selectively to a registered change in source value rather than merely shrinking all channels together.
- Each monitored source had a registered causal replacement cost in both base and shift environments.
- The frozen reference-v2 allocation account outperformed every eligible registered baseline in the tested alpha/beta comparisons.
- These relations survived a preregistered, outcome-embargoed, hash-bound confirmation with fixed seeds, fixed order, fixed gates, fixed margins, and no interim scientific access.

## What this does not establish

- It does not establish that the monitored signals are strict metarepresentations, representational redescriptions, or representations of their generating processes.
- It does not establish consciousness, introspection, or a claim about human cognition, neuroscience, clinical systems, or deployed artificial systems.
- It does not establish architecture-, optimizer-, task-, or domain-general laws.
- It does not yet earn the canon's stronger word **ecology**, which requires reproducible relations among independently validated systems and belongs later in the experimental ladder.
- It does not test or revive excluded R, L, G, Arm R, `eta>0`, U2-T, comparison-fixture, `k=2`, or `k=3` claims.
- Valid-population disconfirmation and equivalence analyses remain conditional and do not identify an unconditional mechanism effect; the frozen selection-bias disclosure remains in force.
- Reference v2's success is limited to the frozen comparisons and is not a claim that a biological or artificial system literally computes the reference optimization.

## Canon correspondence

This wording follows the Sīdus canon:

- E1 asks whether several internal-state monitors placed under scarcity compete predictably.
- The inference order is local before general: define, test locally, replicate, then generalize.
- Useful internal telemetry is not automatically strict metarepresentation.
- “Ecology” is reserved for reproducible relations among independently validated systems, with E5 identified as the first stage where the term can become an empirical result rather than a programme name.

Canonical sources read on the freeze date:

- https://sidus.quest/#/primer/doc-i
- https://sidus.quest/#/primer/doc-ii
- https://sidus.quest/#/results

## Bound identities

The completed RC4 execution package itself remains unmodified. This claim record is bound to its independently verified final identities:

| Record | SHA-256 |
|---|---|
| Confirmation data | `eb837612ef2f1c18fc0ac4816c2553f2a202f0cf99a58807bf8970f6b962087e` |
| Scientific adjudication | `aea55869cb72929d0622c55626740f2033506e3e7e0fd23db5fbd3d01ca5386d` |
| Execution completion | `3ebb5cab32978ac6c50ae1e0e5c1163f2d71b4d3e6739d7a8f4b4cffbf603fba` |
| Ledger | `908e31425c2a1e366dfa4e5338f99c33b5745f7966442b19d9c0d0eef56057a9` |
| Public status | `71b8278923aa26e760049c0943febfc9af687db0c4d8a78c27f0ed676e9f5967` |
| Run lock | `e10e91995929e3737e089267685db9136aee72afbfc8b6ae6ec9eb18610a014b` |
| Frozen contract | `e98f110cd866a9bc89f00ca80730d8927e30e5f5a2204c1e1dc15e2eeda36f42` |
| Frozen manifest | `d522c77a3f6d650063d03504df9ca373b96189429247f4c2ba9854d4e92d8d7e` |
| CF-4 review record | `a478399c0bc84938460f2fa112db3062e5ff75fe3d404b2cf64029edcb5a44a8` |
| Adopted decision | `efdbdaec97a793f9eb9031fb68d4453be851a73b06ce68b85ef1ef0e200cf1ef` |

The final analyzer exited with code 0; the frozen environment verifier passed every check with an empty distribution diff; all final registered hashes matched; and the final hash-chained lean backup was verified after all 3,006 conditions became terminal.

## Freeze rule

This wording may be quoted, translated, or shortened only if its declared scope and exclusions remain intact. Any broader interpretation is a new claim and requires new evidence.