The compact site mark used in the Observatory navigation to return to the AmberContinuum topology.

Computer systems designed to perform tasks associated with inference, learning, generation, planning, or other forms of intelligent behavior. Amber’s work often studies the coupling between these systems and human operators.

Amber Anson’s AI collaborative research lab. Her questions and ideas develop through a cybernetic loop with AI reasoning into models, code, experiments, explanations, and a traceable research record.

A four-layer computational model that moves from planetary phase through resonance and archetypal projection to symbolic output.

A public prototype that looks at a learning situation and suggests what kind of support may help next.

A defined interface through which software components exchange requests and structured results.

Visible rules for translating a modeled pattern into symbolic language. In the Observatory, they support reflection; they are not diagnoses.

A publicly available theoretical preprint proposing a cross-domain account of light, cognition, machine inference, and symbolic collapse.

A way to say what belongs in the question, what affects it, and what sits outside it.

A way to carry an idea into a new setting while naming what must stay true, what changes in translation, and what would show the crossing failed.

The proposed threshold at which latent resonance becomes a selected symbolic state within the QIAA framework.

A PDOT measure representing resonance minus its preregistered threshold; it is used to describe proximity to a modeled collapse boundary.

A prototype diagnostic framework for computational hierarchy, symbolic-pressure detection, and operational AI-state modeling. Its psychological language describes observable behavior rather than proven subjective states.

A seven-level model of computational drives that describes system-specific necessities without borrowing human motivational psychology.

The ability of a relationship to stay honest about its purpose and limits as conditions change.

Amber’s way of studying and shaping relationships among people, AI, and ideas so the work remains clear as it develops.

What needs to stay true so a process can change without losing its purpose.

A relationship in which people, tools, or ideas affect one another. Amber studies what changes through that contact.

A NeurIPS polymer-prediction dataset using physics-inspired descriptors engineered from SMILES for five material-property targets.

The map interface for following relationships among the areas of AmberContinuum.

Amber’s ten-step practice for developing an idea through reciprocal human–AI reasoning: noticing a pattern, engineering feedback between AI architectures, testing disagreement, building an artifact, challenging the result, and preserving the record.

A persistent identifier used to locate a publicly available digital research object even if its hosting address changes.

A developing framework asking whether computation, meaning, and alignment share a mathematical condition for becoming determinate. Its Lean 4 models contain proved results alongside explicit axioms and open physical hypotheses; formal proof does not establish the proposed physical mechanism.

The Observatory’s demonstration model for relative cyclic phase. It is an explanatory model, not a precision geocentric ephemeris.

Pre-digital systems humans developed to encode, transform, preserve, and transmit observations through cyclic calendars, spatial geometry, symbol, story, and collective memory. AmberContinuum studies their operations without treating symbolic interpretation as proof of a physical mechanism.

The standard document language used to structure webpages and standalone browser-based instruments.

Amber’s framework for keeping the full human ability to notice, understand, choose, act, check, correct, and recover as AI becomes part of consequential work.

Places in ordinary human life where a vital reasoning skill can still be practiced without AI, such as navigation, conversation, repair, and sensory judgment.

The technology company whose cloud quantum processors are used in the QRT hardware research record.

A standard astronomical reference epoch centered on January 1, 2000. The Observatory uses J2000 reference phases to anchor its simplified orbital phase model.

A directed process in which distinct AI models are assigned different functions, contradictions are exposed, and only the structure that survives recursive comparison is retained.

A compact operating technique that discloses the entry condition for an AmberContinuum region. A key does not summarize the region; it tells the visitor how to begin thinking inside it.

An annual machine-learning research conference. The polymer dataset in Experimental Programs was created for a NeurIPS prediction challenge.

An exploratory, 12-question interactive assessment about noticing how prompts shape responses. It includes a debrief and is not a validated diagnosis.

An open research platform used to register protocols and preserve research materials. PDOT was registered there on 17 May 2026.

A visual model of how human interaction creates external curvature that can reshape AI inference.

A map of who affects whom, what each side relies on, and where a relationship may become strained.

A registered study of human–AI interaction measures under conversational perturbation. Pilots are documented; no confirmatory outcome is claimed here. It does not test AI consciousness or inner life.

A deliberate or environmental change applied to a system so that its response, limits, and preserved commitments can be observed.

A PDOT measurement family covering detection latency, recovery fidelity, and preservation of the coupling after a perturbation.

A research formulation that treats human–AI interaction as a measurable coupling field rather than locating the dynamic solely in the human, model, or transcript. PSI is the field’s name, not a further acronym.

A proposed cross-domain framework connecting light, cognition, machine inference, and symbolic collapse.

A publicly available theoretical preprint accompanied by a preregistered hardware program that preserves protocols, negative findings, corrections, and replication records.

A preregistered IBM quantum-hardware research program testing predictions associated with Quantum Resonance Theory.

The Zodiac Engine’s geometric organizing scheme for relating archetypes across a 360-degree symbolic field.

A way to see whether a relationship can adapt under pressure or becomes rigid or fragile.

A method for training or aligning AI systems using preference signals derived from human judgments.

The configurations a system is allowed to occupy, including its explicitly named failure states.

A line notation that represents molecular structures as text. It is the source representation used to engineer descriptors in the crystalline-polymer dataset.

A tool for noticing when human–AI work starts to drift, repeat itself, lose clarity, or narrow human choice.

An extension of the STK-Flux kernel across development sessions, handoffs, proxy states, and commit-safety decisions.

A participating form of cognition or computation—human, machine, symbolic, social, or biological—described without assuming they function identically.

A public framework, validator, and set of papers examining information loss and the limits of descriptions of AI behavior. Its interpretation remains open to challenge.

A transparent mapping from modeled state into an interpretive symbolic basis. It produces a reading, not proof of a physical equivalence.

The abbreviated systems-status label used in the map’s lower navigation key.

The part of AmberContinuum where an idea becomes a working tool, gets tested, and changes in response to what the test reveals.

The time standard used by the Observatory to keep selected dates stable across a visitor’s local timezone.

The commitments a process needs to keep as it changes.

An inspectable model of a native human computational system. It separates celestial input, geometric transformation, archetypal mapping, and narrative output so visitors can see where calculation ends and interpretation begins.