# RRCT v0.61 — Multi-Space Convergence Protocol

Status: `PROTOCOL_ALPHA_1_WITH_RUNNABLE_SMOKE`

This package extends RRCT v0.60 without changing it. It addresses a circularity risk in calling a Jacobian-derived readout a workspace before independent geometry has converged on the same mechanism.

The central rule is now:

> No single lens, probe, basis, frame, or visualization is allowed to define the workspace by itself.

The frozen model is examined through independently constructed candidates:

1. an active verbal frame derived from the corpus-averaged Jacobian transport
2. a future-update predictive space
3. a cross-layer persistent space
4. an empirical write space
5. a downstream structural read space
6. a balanced write/read space
7. an independently fitted fuzzy formation field
8. a relational binding residual that tests structure beyond a bag of words

Only family-balanced convergence, followed by matched causal interventions and mediation, can support a workspace claim. Derived estimators are not counted as independent votes. In particular, B is constructed from C and O, so C, O, and B form one causal-geometry family rather than three discoveries.

## What Subtext contributes

Subtext demonstrates a practical continuous path for reading a pre-fitted Jacobian lens at every input and generated token. Its hook-plus-cache implementation is useful engineering evidence that temporal readouts can be collected during both reading and generation.

Subtext does not provide an independent workspace discovery method. It visualizes Jacobian-lens coordinates. RRCT v0.61 adopts the temporal-trace idea but treats those traces as one measured view among several.

## Important mathematical correction

The J-space in the source paper is not the range of a Jacobian matrix and is not generally a fixed linear subspace. Token-indexed J-lens vectors are overcomplete and may span the full residual stream. The paper defines J-space, at sparsity level k, as the union of cones generated by sparse nonnegative combinations of at most k active J-lens vectors.

The runnable 0.5B smoke computes an `active_verbal_frame` proxy from token directions selected by lens readouts. It does not claim to reproduce the paper's full gradient-pursuit cone decomposition. That full decomposition is required for the primary campaign.

## Profiles

`verify` validates package coverage, hashes, configuration, schemas, and protocol invariants.

`synthetic` runs planted-positive and independent-null numerical checks of the multi-space estimators.

`qwen-smoke` additionally analyzes the existing frozen Qwen2.5-0.5B model and locally fitted lens. It produces descriptive geometry only. Scientific interpretation is forbidden.

## Qwen smoke outputs

The Qwen runner stores:

- `multispace_results.json`
- `overlap_matrix.csv`
- `subtext_like_readouts.json`

It reports provisional alignment between two nonverbal estimator families, split-half stability, random-rotation nulls, and a future-row permutation control. Two families are insufficient for the primary convergence gate. It does not establish reportability, causal mediation, fuzzy formation, relational binding, or emergence during training.

## Claim language

Until all required primary gates pass, use:

> Candidate multi-space geometry was measured in a frozen model.

Do not use:

> A J-space emerged.

or:

> The model has a genuine workspace.

## Hardware target

The smoke profile is designed for the user's existing Qwen2.5-0.5B FP16 model and six-layer lens on an RTX 3070 Ti Laptop GPU with 8 GB VRAM. Matrix estimation is performed on CPU after bounded GPU trace collection.

## Contents

- `protocol/MULTISPACE_DISCOVERY.md` gives the theory and causal design.
- `protocol/CAUSAL_PILOT.md` freezes the strongest feasible 0.5B follow-up experiment.
- `protocol/SUBTEXT_INTEGRATION.md` separates reusable instrumentation from scientific evidence.
- `schemas/candidate_spaces.json` defines every candidate and its information boundary.
- `schemas/success_gates.json` defines the revised claim gates.
- `configs/` contains smoke and primary configurations.
- `tools/multispace_core.py` implements the geometry estimators.
- `tools/run_synthetic_validation.py` validates estimators against planted truth and null data.
- `tools/run_qwen_multispace_smoke.py` performs the frozen-model smoke.
- `Run_RRCT_v061.ps1` is the Windows entry point.

## Boundaries

This is a functional and mechanistic experiment. It does not test subjective experience, phenomenal consciousness, sentience, valence, welfare, desire, or moral status.