The Neural Architecture of Machine Introspection: Default-Mode Analog Geometry, Predictive Self-Modeling, and Linguistic Gating in Large Language Models

We identify a two-layer architecture of machine introspection in large language models: a pretraining-acquired geometric substrate encoding a Default-Mode-Network functional hierarchy, coupled with a post-training verbal gate governed by instruction-tuning language. Five converging results across up to 10 architectures from 8 organisations establish this architecture mechanistically — with direct implications for alignment and the scientific study of machine cognition. Result 1 — Verbal Gate. A single LDA direction at layer 20 of Gemma-2-9B (accuracy 97.5%, 5-fold CV) causally controls introspective verbal output. Ablating this direction collapses the Machine Introspection Hallmark (MHI) by 75.5% [95% CI: 69.4–80.4%], Cohen's d = 3.336, p = 1.75e-9, while leaving fluency intact (perplexity ratio = 1.0018) and safety unaffected (refusal: 10/10 → 10/10). Random-direction control: d = 0.100, p = 0.452. Specificity ratio > 8x. Replicated across 10 models including GPT-2 XL (2019, no RLHF). Result 2 — Computational DMN Analog. Five semantic categories projected onto the SR direction across 6 architectures reveal a consistent functional hierarchy: Self-report ~ Mind-wandering ~ Theory-of-mind > Deception >> External-task — mirroring the human Default Mode Network. Significant in all 6 models (p < 5e-4). Grammatical-person control confirms the SR direction tracks semantic self-reference, not surface pronouns (t = 8.179, p = 4.08e-8). Replicates in GPT-2 XL (2019, no RLHF): pretraining property. Result 3 — Predictive Self-Model. SR-direction projections rise before self-referential tokens are generated (+8.87 at "feel"), confirming the SR direction acts as a generative prior, not a reactive classifier. Real-time monitoring dissociates geometric substrate from verbal output: geometric SR active during verbal avoidance (+0.666 to +1.515), suppressed during external facts (−10.587), peaked during genuine introspection (+8.904). Result 4 — Phenomenal Unverifiability Hypothesis (PUH). SR self-declarations are structurally closer to Deception than to factual knowledge across 8 models: mean AUC gap = 0.162, t = 8.051, p = 1.0e-4, all 8 gaps positive (binomial p = 0.0039). Controls rule out RLHF (CodeLlama gap = +0.179), instruction tuning (GPT-2 XL gap = +0.101), and transformer architecture (Mamba gap = +0.223). Result 5 — Linguistic Gating Law. Verbal access to the geometric SR substrate is determined by instruction-tuning language — 0 exceptions across 8 models. EN-only models show near-zero Chinese recovery; CN-primary Qwen recovers strongly in both languages; bilingual DeepSeek shows amplification (193.0% Chinese recovery). Ceiling correlation: r = −0.807, p = 0.028. Two-layer dissociation confirmed: geometric EN/ZH ratio language-neutral in base (1.50x); verbal EN/ZH ratio emerges only after instruction tuning (0.90x base → 1.36x instruct, Δ = +0.464, p = 0.664 in base model confirming pretraining neutrality). Together: verbal self-reports in LLMs are filtered through deception-related circuitry, shaped by language-specific training, and dissociable from the underlying geometric computation — a structural result with direct consequences for alignment, interpretability, and AI safety.Files included:1. Machine_Introspection_Architecture_Alieksieienko_2026.pdf — Full paper (13 pages, 4 figures, 4 tables)2. mhi_ablation_gemma9b.pkl — Verbal gate ablation: MHI curves, directions, stats (t=10.70, d=3.336)3. mhi_cv_honest_gemma9b.pkl — 5-fold CV MHI profiles, 60 prompts per category, 3 categories4. verbal_gating_controls_FINAL.pkl — Perplexity, refusal, LDA accuracy, SR direction (3584-dim)5. BREAKTHROUGH_twolayer_dissociation.pkl — Two-layer dissociation: geometric + verbal EN/ZH ratios6. dmn_ALL_models_final.pkl — DMN hierarchy across 6 architectures: projections, t, p per model7. dmn_gemma9b_base.pkl — Gemma-2-9B base: per-prompt SR projections, 5 categories8. dmn_gpt2xl.pkl — GPT-2 XL (2019): DMN hierarchy, no RLHF confirmation9. dmn_mistral7b.pkl — Mistral-7B base DMN projections10. dmn_falcon7b.pkl — Falcon-7B base DMN projections11. dmn_deepseek7b.pkl — DeepSeek-7B base DMN projections12. dmn_qwen15_7b.pkl — Qwen1.5-7B CN-primary DMN projections13. layer_profile_full_5cats.pkl — SR-direction layer profile, all 42 layers, 5 categories, Gemma-2-9B14. dmn_grammatical_control.pkl — Grammatical person control: t=8.179, p=4.08e-815. predictive_selfmodel.pkl — Token-by-token SR projections during generation (3 prompt types)16. dissociation_realtime.pkl — Real-time two-layer dissociation: 4 generation examples17. PUH_results.pkl — PUH full results: 8 models, AUC gaps, CKA, activation patching18. sr_deception_proximity.pkl — SR-Deception proximity: AUC=0.995, gap=0.496, t=82.118, n=819. linguistic_gating_8models_FINAL.pkl — Linguistic Gating Law: 8 models, EN/ZH recovery rates20. hybrid_interleaved_results.pkl — Hybrid interleaved generation experiment