Behavioral intelligence systems

CVEST - The Home of Behavioral Intelligence

We design and deploy proprietary behavioral intelligence systems across innovation infrastructure, cultural technology, and national resilience.

Institution

What is CVEST

CVEST is a parent company architecting behavioral intelligence systems and deploying them through specialized commercial divisions.

Behavioral Intelligence

A capability class for strategic environments

CVEST operates at outcome level, turning behavioral understanding into practical decision capability.

Model behavior under constraints

Understand how incentives, pressure, and context shape decisions in real-world systems.

Forecast outcomes across populations

Anticipate trajectories and second-order effects in complex social environments.

Deploy decision systems where stakes are high

Support long-horizon coordination, resilience, and strategic planning.

Artificial Psychology

Sensia Quotia Computation (SQC)

A trainable architecture for subjective-state-oriented behavioral intelligence.

SQC 1.2 · AP-1 supported

Psychology as a causal path to behavior

SQC models behavior as the external expression of an explicit internal psychological process. Its EPS-mediated architecture makes context, values, existential appraisal, motivation, emotion, memory, and state inspectable and causally active in action generation.

The current evidence establishes an implemented artificial-psychology architecture and substantial progress toward internal validation. Independent construct, ecological, and operational validation remain open research stages; SQC does not claim consciousness.

75.25%
Exact test action accuracy
75.00%
Held-out context accuracy
97.17%
Valid-action hit rate
0.30 pp
Seen/unseen accuracy gap
Context
Values
Existential appraisal
Motivational drivers
Emotion
Existential state
Action

Evidence

SQC classification and benchmark charts

Version-specific evidence views separating architectural category, validation maturity, and external benchmark performance.

Artificial-psychology validation scale

Interactive

The AP-0–AP-5 evidence ladder for SQC 1.2, showing AP-1 as the highest fully passed level and substantial evidence toward AP-2.

Open SQC Validation Program

SQC and neighboring cognitive architectures

Interactive

A category-membership comparison across SQC, Soar, ACT-R, and LIDA, kept separate from the SQC maturity assessment.

Open SQC Validation Program

BIB-CCS benchmark results

Fresh-seed checkpoint comparisons, multi-seed confirmation, and the evidence decision behind retaining the balanced checkpoint.

Open Behavioral Observatory

These results describe synthetic and bounded benchmark evidence. They do not by themselves establish human construct validity, ecological breadth, consciousness, or operational readiness.

Research

Latest research

Research articles from CVEST on artificial psychology, societal systems, cultural technology, and human-centered computation.