We write papers about everything we build — not because academia requires it, but because papers force honest thinking. Every claim needs a hypothesis. Every hypothesis needs an experiment. Every experiment needs results reported as they are, not as we wished they were.
What We've Demonstrated
These are the capabilities we've validated with experimental data. Each links to a paper with full methodology, results, and limitations.
Learning
| Capability |
Result |
Paper |
| Knowledge extraction from documents |
95%+ provenance attribution, 1.8M+ triples |
102 |
| Continuous learning during conversation |
97.6% extraction precision, +33.1% graph growth in 8hr |
104 |
| Knowledge transfer between beings |
Equal quality, 100,000x faster than corpus learning |
113 |
Reasoning
| Capability |
Result |
Paper |
| Symbolic-path query speed |
11.3ms (1,345x improvement over baseline) |
103 |
| Hallucination on grounded facts |
0% (provenance-tracked queries) |
103, 108 |
| Unified cognitive architecture |
9/9 hypotheses validated, 82% task autonomy |
108 |
| Episodic memory for mistake prevention |
85.7% mistake prevention rate |
107 |
| Safety-critical confidence routing |
0 safety violations, 4/4 hypotheses validated |
114 |
Knowledge Organization
| Capability |
Result |
Paper |
| 7-layer semantic architecture (Y0-Y6) |
5/5 hypotheses validated |
117 |
| Predictive memory loading |
76.2% next-topic prediction accuracy |
110 |
| Fractal knowledge indexing |
0.003ms lookup, <100ms startup |
112 |
| Procedural skill transfer |
90.5% skill transfer between beings |
119 |
| Metacognitive calibration |
3.1% expected calibration error |
120 |
Evaluation
| Capability |
Result |
Paper |
| AI understanding certification |
9-dimension framework, 66 measures, 6 levels |
122 |
| V11→V12 architecture comparison |
+8.7 to +15.8 ACF aggregate improvement, 3/4 beings level up |
122 |
| Parallel signal voting (V12) |
8/8 hypotheses validated, 252 tests passing |
124 |
All Papers
| # |
Title |
Status |
| 102 |
Training Neurosymbolic Beings |
Ready |
| 103 |
Reasoning Performance |
Ready |
| 104 |
Continuous Learning (Crystallization) |
Submitting to NAI Journal |
| 107 |
Graph-Native Episodic Memory |
Ready |
| 108 |
The Perception Brain |
Ready for review |
| 110 |
Semantic Field of View |
Draft complete |
| 112 |
Fractal Knowledge Loading |
Draft complete |
| 113 |
COG Transfer Experiment |
Submitting to NeSy 2026 |
| 114 |
Domain-Specific Confidence |
Validated (4/4) |
| 117 |
Yurtle Semantic Graph Architecture |
Draft complete |
| 118 |
Predictive Processing |
Partial validation (6/13 hypotheses) |
| 119 |
Y5 Procedural Memory |
Partial validation (2/5) |
| 120 |
Y6 Metacognition |
Partial validation (2/5) |
| 121 |
Wikidata Knowledge Enrichment |
Awaiting experiments |
| 122 |
AGI Certification Framework |
Spec complete, V11→V12 comparison data collected |
| 124 |
Cognitive Signal Fusion (V12) |
Validated |
Papers 101, 103 have been folded into Paper 108 (the unified architecture paper). Papers 105, 106, 115 are superseded or internal. Papers 109, 111, 116 are on hold pending implementation work.
Research Integrity
- Simulation results labeled as simulation. Live results labeled as live. When there's a gap, we analyze it.
- Failures reported alongside successes. Paper 118 missed its target. Paper 119's speed hypothesis didn't hold. We publish the skeleton alongside the marlin.
- Limitations sections are thorough. Not two sentences at the end. Numbered, specific, honest.
- Reproducibility. Code, data, and experiment designs are maintained in the NuSy repository.
Contact: [email protected] | ORCID: 0000-0002-3912-0080