02
Cognitive Models

A model series designed to operate inside a cognitive architecture.

Ingenious models are purpose-built interfaces and cognitive components. Deep capability comes from the complete system: CFAM, DFAM, agents, memory, retrieval, knowledge graphs, tools and verification.

Interactive explorer

Explore the model family.

Select a model to review its role, operating scale, deployment profile and connection to the wider cognitive system.

Micro Language Interface

ING-M-LI-v1

A compact language interface for edge, IoT and embedded cognitive terminals. It parses commands and routes structured intent without depending on external APIs.

Parameters~1M
Context512 tokens
PrecisionINT4 / INT8
DeploymentCPU / embedded
  • Real-time intent parsing
  • Structured command decomposition
  • Offline-first deployment
  • Sub-10 ms target latency on embedded CPU
  • Designed for constrained devices
  • No external API dependency
CFAM

Concept Formation and Appraisal Model.

CFAM produces structured conceptual intelligence through explicit stages rather than relying only on statistically likely language.

01Input Concept Layer

Frames the request, domain, intent, constraints and contextual conditions.

02Concept Extraction

Identifies concepts, entities, assumptions, goals, tensions and hidden dependencies.

03Principle Identification

Selects ethical, logical, scientific, governance and contextual principles.

04Reference Validation

Checks concepts against approved evidence and source constraints.

05Relationship Discovery

Builds principle-guided connections between concepts and systems.

06Derived Concept Generation

Produces scored new concepts, frameworks and possible solutions.

07Architecture Formation

Organises concepts into coherent structures, workflows or system designs.

08Concept Appraisal

Evaluates quality, coherence, originality, risk and operational usefulness.

DFAM

Decision Formation and Appraisal Model.

DFAM transforms complex problems into explicit decision structures, scenario analysis and auditable multi-dimensional scoring.

Evaluation dimensionWhat it examinesExample output
EthicalHarm, dignity, consent, fairness and responsibility.Ethical score and blocking risks.
ScientificEvidence quality, assumptions and empirical validity.Evidence confidence and missing proof.
SocialStakeholder impact, access, equity and cohesion.Impact map and affected groups.
GovernancePolicy, regulation, accountability and auditability.Controls and approval requirements.
EconomicCost, benefit, resources, efficiency and sustainability.Cost-benefit and implementation score.
Architecture

The model is the interface. The system is the intelligence.

Knowledge remains dynamic, memory preserves continuity, agents execute work, CFAM forms concepts, DFAM appraises decisions and verification remains mandatory.

#FunctionSystem locationUpdate modeWhy it is explicit
01LanguageLI modelsVersioned

Communication and intent parsing.

02ConceptsCFAMStructured process

Traceable concept formation.

03DecisionsDFAMScored appraisal

Auditable recommendations.

04KnowledgeRAG + graphInference-time

Current evidence without full retraining.

05ContinuityMemoryPersistent context

Adaptation across tasks and sessions.

06ReliabilityVerifierEvery output

Logic, fact and consistency checks.

Own your AI future.

Build, customize and deploy cognitive AI with complete control.

Demo accessBook a walkthrough