Research archive
Research
Every essay, tutorial, and experiment in the open — the working notes behind computational sovereignty.
12 posts tagged knowledge-graphs clear
AI 45sovereign-ai 36Ollama 35ai-agents 31local-first 24local-first-ai 22Python 21RAG 21compile-time-ai 20knowledge-compiler 19knowledge-graph 19LLM 19Tutorial 19ollama 16Local LLMs 15Machine Learning 15Next.js 15FastAPI 13knowledge-graphs 12local-ai 12AI Agents 11MCP 11local AI 10rag 10AI Development 9OpenAI 9React 9Knowledge Graphs 8open-source 8sovereign AI 8
- Jul 26, 2026Does the Compiler Know When It's Wrong? Semantic Rediscovery and the Limits of Falsification-AwarenessPopper-ORKG ended with two honest failures: temporal rediscovery was near-zero, and the compiler confidently rebuilt Lamarckian inheritance. v4 exists to ask whether those are fixable weaknesses or fundamental limits — by adding measurement, not by tuning the scorer. The answer turned out to be one of each. Mechanism-level semantic matching shows rediscovery was a measurement artifact all along (0.63 vs 0.005 exact-match, flat across temporal gaps). But a falsification-awareness consensus check could not downgrade Lamarckism or caloric: the ORKG index contains no detectable refutation edges, so the near-miss is a genuine limit of the knowledge substrate. Everything in popper/ and orkg/ stayed byte-for-byte frozen, md5-verified. Run locally on a 35B model.ai-agentscompile-time-aiknowledge-compilerlocal-first-aiscientific-discoveryfalsifiability
- Jul 25, 2026Does the Hypothesis Compiler Generalize? Running Popper on a Corpus I Did Not BuildPopper beat its controls on five hand-curated benchmark literatures — but I built those corpora, so the result could be an artifact of clean data. This experiment removes that objection: the identical pipeline, byte-for-byte, run on a corpus I did not build — the Open Research Knowledge Graph, 6.3 million triples across 65,689 papers and seven scientific disciplines. Same schema, same rubric, same four-way controls; the only independent variable is the corpus. The compiler still wins (0.85 composite vs 0.56 keyword vs 0.00 ungrounded), an ablation proves the architecture is load-bearing (remove graph reasoning and it collapses to zero), and the grounding audit finds zero hallucinated citations across 265 evidence items. It also fails honestly: exact temporal rediscovery is near-zero, and it confidently rebuilds Lamarckian inheritance. Run locally on a 35B model.ai-agentscompile-time-aiknowledge-compilerlocal-first-aiscientific-discoveryfalsifiability
- Jul 14, 2026The Recursive Research Compiler: Turning Compile-Time AI Inward with the Knowledge Compiler SDKA case study in Compile-Time AI: how the Knowledge Compiler SDK turns Markdown into inspectable intermediate representations, and how pairing it with an agent like Hermes lets a codebase compile its own next generation of passes.compile-time-aisovereign-aiknowledge-graphslocal-firstagentscompilers
- Jul 11, 2026Knowledge Compiler: Why I'm Building a Compiler for Human Knowledge Instead of Another RAG SystemKnowledge Compiler transforms collections of Markdown documents into statically-deployable semantic artifacts — knowledge graphs, concept hierarchies, vector embeddings, and cluster maps — using a multi-pass compilation pipeline. No runtime LLM inference required.knowledge-compilerknowledge-graphsRAGGraphRAGAI-infrastructurelocal-ai
- Jul 10, 2026The Telemetry Intelligence Engine: A Local-First GraphRAG System for Website AnalyticsA spec-driven walkthrough of the Telemetry Intelligence Engine (TIE): a local-first GraphRAG system that turns GA4 telemetry and site content into a behavioral knowledge graph an operator can query in natural language.graphraglocal-first-aiknowledge-graphsanalyticssovereign-ai
- Jul 5, 2026Retrieval Architecture: Memory Systems That CompoundMemory systems and retrieval architecture for sovereign AI. Sovereign Memory Bank, Dynamic Persona MoE RAG, Objective05, and GraphRAG — the subsystems that make retrieval compound over time.retrieval-augmented-generationsovereign-memory-bankdynamic-persona-moe-ragobjective05sovereigntyspecgraphrag
- Jul 5, 2026Sovereign AI Architecture: Building Compounding IntelligenceA comprehensive synthesis of four years of architectural investigation into sovereign AI. Ties together the Sovereign Intelligence Stack, Sovereign Memory Bank, Dynamic Persona MoE RAG, Objective05, and SovereignSpec into one unified compounding intelligence architecture.sovereign-aiai-architecturecompounding-intelligencerecipe-compilersignal-routerevaluation-loop
- Jul 4, 2026The Sovereign Intelligence Stack: Building Compounding AI InfrastructureBuilding a 5-layer architecture where every AI decision compounds into the next layer. The recipe compiler, signal router, autonomous evaluation loop, and more — with working code.sovereign-intelligenceai-infrastructurelocal-firstagent-recipesknowledge-graphsautonomous-evaluation
- Jul 3, 2026The Model Is Not the Product: Residual State, Compiled Agents, and Optimization LoopsThree converging research threads — Apple's Residual Context Diffusion, LMSYS/SGLang agentic execution graphs, and constrained optimization for agent loops — collapse into a single architectural claim: the model is no longer the product. The loop is. With code examples, arxiv citations, and cross-references to Objective05, Sovereign Memory Bank, and Dynamic Persona MoE RAG.model-is-not-the-productresidual-context-diffusionsglangexecution-graphsconstrained-optimizationagent-loops
- Jul 3, 2026The Loop Is the Product: Inside the Sovereign Intelligence ObservatoryA technical deep dive into the Sovereign Intelligence Observatory: a six-component, local-first pipeline that turns every agent run into a versioned recipe, routes evaluation by confidence tier, detects capability drift with KS and PSI statistics, extracts tacit expert judgment into decision trees, and manages the supervised-to-autonomous transition. With architecture diagrams, code from the actual repo, and the case for why observability -- not the model -- is the real product.sovereign-intelligence-observatoryagent-recipesdrift-detectionexpert-signal-routingtacit-knowledge-extractionautonomy-ladders
- Jan 16, 2026From Fragmented Experiments to Cognitive Synthesis : The Evolution of SimulacraHow two years of AI experimentation—from basic chatbots to autonomous architectures—converged into Simulacra, a living cognitive system that synthesizes knowledge, memory, and personality into something profoundly new.AIautonomous-agentsknowledge-graphspersona-engineeringdigital-resurrectionsimulacra
- Jan 11, 2026Memory Preservation Invariants: A New Class of Autonomous Agent ArchitecturesFormalizing memory preservation constraints in AI systems to enable identity reconstruction and long-horizon autonomy beyond traditional RAG and agent frameworks.AIautonomous-agentsknowledge-graphsmemory-preservationdeterministic-pipelinescomputational-sovereignty