Ontology-Amplified Distillation and Contextuality Auditing for Sovereign Enterprise Language Models: A Combined Proof-of-Mechanism and Negative-Results Method Study
Quick Answer
This study examines ontology-amplified distillation using a Qwen3.6-27B model, achieving a 90% grounding rate on Vietnamese financial tasks, but lacks evidence for superiority over GPT-5.
Quick Take
This study examines ontology-amplified distillation using a Qwen3.6-27B model, achieving a 90% grounding rate on Vietnamese financial tasks, but lacks evidence for superiority over GPT-5. It also introduces a contextuality-audit method, revealing no residual contextuality in enterprise-agent routing, which questions the model's deployability and safety.
Key Points
- Qwen3.6-27B model adapted with ontology-grounded .
- Achieved 90% grounding rate on 40 Vietnamese financial tasks, equal to GPT-5.
- Study lacks evidence for model superiority or deployability.
- Contextuality-audit method shows zero residual contextuality in enterprise-agent routing.
- Findings suggest need for governance diagnostics in model deployment.
Paper Resources
📖 Reader Mode
~2 min readAbstract:Regulated financial institutions operating under data-residency rules need tenant-owned language models that can run inside the institution's perimeter. This paper combines two related FAOS studies into one mechanism-and-control article. First, it reports a reduced-power proof-of-mechanism study of ontology-amplified distillation: a Qwen3.6-27B student is adapted to the Foundation AgenticOS ontology through supervised fine-tuning on frontier-teacher trajectories and ontology-grounded direct preference optimization (DPO), trained locally on a single Apple M5 Max from 47 synthetic, English-language, cross-domain preference pairs. On 40 held-out Vietnamese financial-domain tasks, the distilled student grounds 36 of 40 tasks (grounded rate 0.90; mean ontology term-coverage r_onto = 0.95 on a metric floored at 0.50), equal to the GPT-5 frontier baseline, which also grounds 36 of 40. The outcome is underpowered to establish equivalence: the paired-difference 95% confidence interval spans +/-4 tasks, and the run does not test or show the pre-registered amplification prediction that the student should exceed the frontier. Second, the paper consolidates a contextuality-audit method for enterprise-agent routing. In a separate negative-results pilot, the corrected canonical Contextuality-by-Default degree is zero for all Phase 1.3 groups in both the local-Qwen run and an explicitly labeled Gemma replication check; the useful signal is direct influence and construct coupling, not surviving residual contextuality. Together, the studies pair an ontology-grounded model-building mechanism with a governance diagnostic for deciding when apparent disagreement should trigger prompt standardization, multi-agent synthesis, or human review. The evidence supports neither deployability, safety, superiority, statistical equivalence, nor a contextuality-positive routing rule.
| Comments: | 15 pages, 2 figures. Combined proof-of-mechanism and negative-results method article consolidating ontology-amplified distillation with contextuality-audit routing for enterprise agents |
| Subjects: | Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG); Multiagent Systems (cs.MA) |
| Cite as: | arXiv:2607.11948 [cs.AI] |
| (or arXiv:2607.11948v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2607.11948 arXiv-issued DOI via DataCite |
Submission history
From: Thanh Luong Tuan [view email]
[v1]
Sat, 11 Jul 2026 15:42:40 UTC (107 KB)
— Originally published at arxiv.org
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.AI
See more →The Emerging Paradigm of Geospatial Foundation Models: From Pre-Training to Agentic Reasoning
The paper introduces Geospatial Foundation Models (GeoFMs), AI/ML models pre-trained on extensive geospatial datasets, enabling domain experts to fine-tune them for specific tasks. This paradigm shift democratizes access to advanced AI/ML while ensuring security, and proposes a framework for cost-effective adaptation strategies. The vision of Agentic Geospatial Reasoning is also presented, where Large Language Models orchestrate GeoFMs to automate complex analytical workflows.