Experiment estate
Executable experiments and generated artifacts exist across multiple mathematical and computational research tracks.
HBIF Research ProgrammeThe interesting question is not whether a pattern exists.
The interesting question is not whether a pattern exists.
It is whether the pattern survives.
Polynovea's Human Behavioural Intelligence Framework is an evolving research architecture built from real domain systems, executable experiments, behavioural-state modelling and cross-domain tests. It is not presented as a finished universal theory of human behaviour.
Public evidence ledgerHBIF / 2026.09.19
Evidence state
What exists today - and what still has to be proved.
Uncertainty is part of the result. The ledger separates implemented work, operationalised research, substantial domain evidence and claims that remain unestablished.
BPM
A subset of the research has been turned into a working behavioural state-and-dynamics pipeline.
Hospitality corpus
A large historical domain corpus and product/research lineage provide real empirical material rather than synthetic theory alone.
Cross-domain universality
Polynovea does not claim that one shared behavioural model has already been validated across independent domains.
Public research map
Selected research tracks.
These are the areas that can be discussed publicly without exposing restricted internal research mechanics.
Real systems before universal claims
HBIF begins with domain-specific systems, ontologies, observations, interventions and outcomes. Hospitality produced the first substantial empirical lineage; Workplace now has Infrakinetic as its Layer 1 operating product.
Mathematics and dynamics implemented against data
The research estate includes implemented statistical, information-theoretic and dynamical-systems experiments with preserved outputs. The point is not mathematical decoration; candidate structures must survive empirical testing.
Behavioral Phase Model
BPM operationalises a subset of the research into a state-and-dynamics pipeline. It is the current Layer 2 technical reference, while shared cross-domain validity remains an open research problem.
Transfer is tested, not assumed
The long-horizon objective is to test whether independently observed domain phenomena converge toward useful shared behavioural structure. Replication, falsification, uncertainty and boundary conditions are required before promotion.
Reasoning systems consume evidence; they do not rewrite it
Future AI and agent interfaces should preserve HBIF observation, state estimate, uncertainty and competing hypotheses separately from an LLM's interpretation or recommendation.
Evidence state is part of the architecture
Implemented, calibrated, hypothesised, restricted and unproven are not interchangeable labels. Public claims are intentionally narrower than the full internal research ledger.
Long-horizon path
From domain evidence to shared mechanisms - only if the evidence survives.
01Real domain systems
02Domain-specific ontologies and observations
03Domain behavioural state and dynamics
04Cross-domain comparison
05Candidate shared abstractions
06Replication and falsification across independent domains
07Domain-independent mechanisms where evidence survives
08Bounded use by AI, agents and decision systems
Claim boundary
What this page deliberately does not claim.
HBIF has solved human behaviour.
HBIF is a universal mathematical theory.
One successful domain proves another domain will behave the same way.
Every implemented research result is a product feature.
Every product customer automatically contributes unrestricted research data.
An LLM can turn a probabilistic HBIF hypothesis into fact by expressing it fluently.
Participatory research
Nightlife behaviour study.
The original 16-question nightlife study remains available as a voluntary Hospitality-domain research input. It now lives under the research programme rather than defining the entire Research page.