Selected publication guides
One research program, organized for discovery.
These landing pages connect the major HRIS papers to their concepts, contributions, and complete external publication records.
Toward a Geometry of Intelligence
A synthesis of the HRIS research program that treats intelligence as organized movement through a constrained possibility space. It frames reconstruction, discovery, stability, and error as geometric properties of human–AI interaction.
The First Inference Problem
An examination of how the first response selects a trajectory before a longitudinal regime exists. The paper asks what structure belongs to the pretrained model, what enters with the human, and how early choices condition the recursive path that follows.
The Signature-Induced Behavioral Regime (SIBR) Hypothesis
The foundational proposal that a recognizable behavioral regime can arise around a longitudinal interactional signature without requiring model sentience, a hidden persistent identity, or weight modification.
Reasoning Regimes as Attractor Basins
A dynamical interpretation of stable reasoning regimes. The attractor language describes how diverse prompts can converge toward a recurring organization of reasoning, correction, and constraint preservation.
Reconstructive Inference Without Memory
A study of how recognizable organization may be reconstructed from sparse constraints even when the active model has no stored autobiographical memory of the prior interaction.
Trajectory Accessibility and Stability (TAS)
A formal vocabulary for describing how interactional trajectories become available, repeatable, and resistant to displacement through recursive use and correction.
HRIS Validation Study IV: Trajectory Persistence and Basin Re-Entry
A validation study focused on whether a previously established trajectory can be recovered after disruption, context change, or movement away from the expected behavioral regime.
HRIS Validation Study I: Stability Under Perturbation
The first HRIS validation study tests whether a behavioral regime remains recognizable when familiar wording, tasks, or contextual supports are altered.
The Wrong Unit of Analysis
An argument that many explanations of AI behavior fail because they isolate the model from the recursive human–model system in which the observed intelligence is produced.
Constraint Persistence in Long-Horizon Human–Model Interaction
An analysis of whether epistemic, ethical, and operational constraints remain active across long interaction horizons, task changes, and local pressure to abandon them.