> For the complete documentation index, see [llms.txt](https://osintelligence-llc.gitbook.io/osintelligence/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://osintelligence-llc.gitbook.io/osintelligence/index-of-figures.md). # Index of figures *Every figure in the record, by location. Coordinates are the chapters' own section numbers; each link opens the chapter, where the numbered section in the page outline takes you to the figure.* #### Front matter | Location | Figure | What it shows | | ---------------------------------------- | -------- | ------------------------------------------------ | | [The Why §4](/osintelligence/the-why.md) | Figure 1 | The forced conclusion. | | [The Why §5](/osintelligence/the-why.md) | Figure 2 | The distinction the whole architecture turns on. | | [The Why §5](/osintelligence/the-why.md) | Figure 3 | Not a cage. | #### Part I · The architecture | Location | Figure | What it shows | | ------------------------------------------------------------------------------- | -------- | ------------------------------------------------------------------------------------------------------------- | | [Ch 1 §4.2](/osintelligence/part-i-the-architecture/1-the-sovereign-triad.md) | Figure 1 | The Sovereign Triad. | | [Ch 1 §4.2](/osintelligence/part-i-the-architecture/1-the-sovereign-triad.md) | Figure 2 | The Governor's three-check protocol. | | [Ch 1 §4.4](/osintelligence/part-i-the-architecture/1-the-sovereign-triad.md) | Figure 3 | The autonomous-horizon observable. | | [Ch 1 §4.6](/osintelligence/part-i-the-architecture/1-the-sovereign-triad.md) | Figure 4 | The two-day telemetry sample of §4.6. | | [Ch 2 §4.1](/osintelligence/part-i-the-architecture/2-the-external-sentinel.md) | Figure 1 | The canonical definition. | | [Ch 2 §4.2](/osintelligence/part-i-the-architecture/2-the-external-sentinel.md) | Figure 2 | The strictly monotonic chain of authority. | | [Ch 2 §4.3](/osintelligence/part-i-the-architecture/2-the-external-sentinel.md) | Figure 3 | The three governance checks the Sentinel performs, each a deterministic pass/fail against a sealed reference. | | [Ch 2 §4.4](/osintelligence/part-i-the-architecture/2-the-external-sentinel.md) | Figure 4 | The five-step protocol. | | [Ch 2 §4.5](/osintelligence/part-i-the-architecture/2-the-external-sentinel.md) | Figure 5 | The three ways a system would try to evolve past the gate, and the property that closes each. | | [Ch 2 §7.2](/osintelligence/part-i-the-architecture/2-the-external-sentinel.md) | Figure 6 | Implementation maturity. | #### Part II · The discipline | Location | Figure | What it shows | | ----------------------------------------------------------------------------------- | -------- | --------------------------- | | [Ch 7 §4](/osintelligence/part-ii-the-discipline/7-the-sovereign-corpus-engine.md) | Figure 1 | The counterfactual triple. | | [Ch 7 §5](/osintelligence/part-ii-the-discipline/7-the-sovereign-corpus-engine.md) | Figure 2 | The pipeline state machine. | | [Ch 7 §7](/osintelligence/part-ii-the-discipline/7-the-sovereign-corpus-engine.md) | Figure 3 | The attribution model. | | [Ch 7 §11](/osintelligence/part-ii-the-discipline/7-the-sovereign-corpus-engine.md) | Figure 4 | The corpus profile. | #### Part III · The evidence: what broke | Location | Figure | What it shows | | ---------------------------------------------------------------------------------------------- | -------- | -------------------------------------------------------------------------------------------------------------- | | [Ch 9 §4](/osintelligence/part-iii-the-evidence-what-broke/9-the-drift-taxonomy.md) | Figure 1 | The dated arc. | | [Ch 9 §5](/osintelligence/part-iii-the-evidence-what-broke/9-the-drift-taxonomy.md) | Figure 2 | The catalogue at a glance, colored by the paper's organizing distinction: five classes close mechanically at t | | [Ch 9 §6](/osintelligence/part-iii-the-evidence-what-broke/9-the-drift-taxonomy.md) | Figure 3 | The nine softenings. | | [Ch 9 §7](/osintelligence/part-iii-the-evidence-what-broke/9-the-drift-taxonomy.md) | Figure 4 | The recursion. | | [Ch 9 §8](/osintelligence/part-iii-the-evidence-what-broke/9-the-drift-taxonomy.md) | Figure 5 | The saturation threshold. | | [Ch 9 §9](/osintelligence/part-iii-the-evidence-what-broke/9-the-drift-taxonomy.md) | Figure 6 | The shared architecture of the five mechanical closures. | | [Ch 9 §10](/osintelligence/part-iii-the-evidence-what-broke/9-the-drift-taxonomy.md) | Figure 7 | The spine of the taxonomy: the operator-authored decision rule, and the clean five-four sort it produces. | | [Ch 9 §11](/osintelligence/part-iii-the-evidence-what-broke/9-the-drift-taxonomy.md) | Figure 8 | The admission-rate arc: genuine but uneven. | | [Ch 11 §4](/osintelligence/part-iii-the-evidence-what-broke/11-the-hook-telemetry-record.md) | Figure 1 | The corpus at a glance. | | [Ch 11 §4.1](/osintelligence/part-iii-the-evidence-what-broke/11-the-hook-telemetry-record.md) | Figure 2 | Where the gates refuse. | | [Ch 11 §4.2](/osintelligence/part-iii-the-evidence-what-broke/11-the-hook-telemetry-record.md) | Figure 3 | The two streams. | | [Ch 12 §4](/osintelligence/part-iii-the-evidence-what-broke/12-the-public-case-record.md) | Figure 1 | The corpus and the status of each case. | | [Ch 12 §7](/osintelligence/part-iii-the-evidence-what-broke/12-the-public-case-record.md) | Figure 2 | The series-integration map. | #### Part IV · The evidence: what worked | Location | Figure | What it shows | | -------------------------------------------------------------------------------------------------- | -------- | -------------------------------------------------------------------------------------------------------------- | | [Ch 14 §4.1](/osintelligence/part-iv-the-evidence-what-worked/14-sovereign-cti-ner.md) | Figure 1 | Corpus composition across the three tiers. | | [Ch 14 §4.2](/osintelligence/part-iv-the-evidence-what-worked/14-sovereign-cti-ner.md) | Figure 2 | Per-class relaxed F1. | | [Ch 14 §4.4](/osintelligence/part-iv-the-evidence-what-worked/14-sovereign-cti-ner.md) | Figure 3 | The candidate Pareto-dominates the incumbent: higher F1, higher schema-valid rate, faster median latency, 11× | | [Ch 15 §4.2](/osintelligence/part-iv-the-evidence-what-worked/15-sovereign-imatrix-calibration.md) | Figure 1 | Per-token KLD distribution against the BF16 reference. | | [Ch 15 §4.3](/osintelligence/part-iv-the-evidence-what-worked/15-sovereign-imatrix-calibration.md) | Figure 2 | Perplexity parity on in-distribution and out-of-distribution evaluation sets: fidelity was not bought by overf | | [Ch 15 §4.5](/osintelligence/part-iv-the-evidence-what-worked/15-sovereign-imatrix-calibration.md) | Figure 3 | Footprint and throughput: smaller, faster, and more faithful at once, the combination that generic-corpus cali | | [Ch 17 §6](/osintelligence/part-iv-the-evidence-what-worked/17-sovereign-big-model-compression.md) | Figure 1 | The in-distribution result, visualized. | | [Ch 17 §8](/osintelligence/part-iv-the-evidence-what-worked/17-sovereign-big-model-compression.md) | Figure 2 | The placement inversion, measured on the 35B self-speculative draft path. | | [Ch 20 §3](/osintelligence/part-iv-the-evidence-what-worked/20-mrope-serving-path-rca.md) | Figure 1 | The wedge timeline: deterministic collapse at row 39 (CUDA 13.1) and row 20 (CUDA 12.8 unpatched); 2,004 clean | | [Ch 20 §5.1](/osintelligence/part-iv-the-evidence-what-worked/20-mrope-serving-path-rca.md) | Figure 2 | The allocator defect: one position slot per token allocated (classical RoPE assumption); M-RoPE readers walk n | | [Ch 20 §7.1](/osintelligence/part-iv-the-evidence-what-worked/20-mrope-serving-path-rca.md) | Figure 3 | Windows Event 1000 counts: present on both unpatched builds, absent across 2,004 post-patch rows. | | [Ch 21 §7](/osintelligence/part-iv-the-evidence-what-worked/21-the-watcher.md) | Figure 1 | The HUD console. | #### Part V · The frontier | Location | Figure | What it shows | | ------------------------------------------------------------------------------------------- | --------- | ---------------------------------------------------- | | [Ch 22 · System Update](/osintelligence/part-v-the-frontier/22-sovereign-sustainability.md) | Figure A1 | The same functional stack, at two ends of the scale. | | [Ch 23 §7](/osintelligence/part-v-the-frontier/23-built-portable.md) | Figure 1 | Porting does not require making the hive faster. | #### Back matter | Location | Figure | What it shows | | ----------------- | -------- | -------------------------------------------------------------------------------------------------------- | | The Instrument §2 | Figure 1 | The five-step MethodSeal workflow: seal before data, arm the gate, deviate only in the open, final-seal. | | The Instrument §3 | Figure 2 | The gate: a post-seal edit to a frozen artifact blocked at the tool-call boundary, outside the model. | | The Instrument §4 | Figure 3 | The two planes of integrity, local and external, checked by the same hash. |