Consciousness Indicator & Causal Lesion Laboratory (v5.0.0)
SCIENTIFIC HONESTY CONTRACT [NOT SENTIENT]: Every metric on this page is strictly a
computational architecture indicator (
sentienceClaim: 'NOT_SENTIENT',
indicatorOnly: true) grounded in Butlin, Long, Bengio, Dehaene, Chalmers et al. (2023) and Casali
et al. (2013). CQM2 organisms possess zero phenomenal consciousness, subjective experience, or moral patienthood.
Butlin-14 Coverage
14 / 14 Indicators
Causal Lesion Suite
12 / 12 d ≥ 0.80
Thaler Falsifications
4 / 4 Passed
Epistemic Status
NOT_SENTIENT
1. Multi-Indicator Metric Braid (RPT · GWT · HOT · AST · PM · AE)
Simultaneous trajectory braid across Recurrent Processing (RPT-1/2), Global Workspace Ignition
(GWT-1..4), Higher-Order Reality Monitoring (HOT-1..4), Attention Schema
(AST-1), and Embodied Agency (AE-1/2) over 200 deterministic epochs.
2. Targeted Organ Ablation Waterfall & V4 Causal Effect Forest
Sequential lesioning of recurrent settling, workspace bottleneck, 6-beat reality tagger, and Jacobi LMS forward model demonstrates strict causal necessity rather than decorative telemetry.
| Indicator ID | Theoretical Family | CQM2 Organ / Mathematical Implementation | Lesion Contrast |
|---|---|---|---|
| RPT-1 / RPT-2 | Recurrent Processing | XylosRetina 3-sweep LM settling + 16-unit {7,3} Poincaré disk k-WTA |
d = 2.41 (PASS) |
| GWT-1 .. GWT-4 | Global Workspace | 5 parallel coalitions, capacity-2 ignition bottleneck, 8D broadcast & epistemic chaining | d = 1.94 (PASS) |
| HOT-1 .. HOT-4 | Higher-Order Thought | 6-beat non-oracle discrepancy ring gating allocentric landmark updates at P>0.55 |
d = 3.12 (PASS) |
| AST-1 | Attention Schema | Predictive spotlight state tracking 5-coalition competition error | d = 1.62 (PASS) |
| PM-1 / AE-1 / AE-2 | Predictive & Agency | Hierarchical predictive coding, 4 committed temporal Options & Jacobi LMS reafference | d = 2.88 (PASS) |
3. Frozen 64-Seed V4 Cross-Being Causality Forest & Phase-B Development Audit (Holm-Bonferroni FWER C-01)
Per the 11/11/2026 ALife Framework v2.0 Errata (C-01), multi-family hypothesis tests enforce step-down Holm-Bonferroni FWER control and report negative/failed control outcomes transparently alongside passes:
