Mechanism
The attention field is a cognitive domain field that governs which other cognitive domain fields are consulted, and to what depth, for a given mutation evaluation. It rests on a structural premise: attention is a finite computational resource, and not all mutations require full-depth evaluation across all cognitive domains. Rather than running every proposed mutation through the complete battery of empathy projection, integrity impact analysis, confidence computation, capability confirmation, and forecasting, the agent allocates evaluation effort according to the stakes of the mutation under consideration. The attention field is the structure that records and enforces that allocation.
Concretely, the attention field determines cognitive engagement depth: a scalar or vector quantity that specifies, for each cognitive domain field, the evaluation depth allocated to that domain for the current mutation evaluation cycle. The field does not perform the evaluations itself; it controls how the other domains are invoked. A mutation evaluated under a shallow attention allocation consults fewer domains at lesser depth; a mutation evaluated under a deep allocation consults more domains, including domains that produce the most expensive analyses, at greater depth.
Stakes-Based Engagement Depth
The disclosed mechanism differentiates engagement depth by the projected impact of the mutation. A low-stakes mutation, one whose projected impact on the agent's state is bounded within a policy-defined low-impact threshold, may engage only confidence and capability evaluation. Under that shallow allocation the agent bypasses full empathy projection, integrity impact analysis, and forecasting engine invocation, because the bounded impact of the mutation does not warrant the cost of those deeper analyses.
A high-stakes mutation is treated differently. A mutation whose projected impact exceeds a policy-defined high-impact threshold, or whose domain-specific characteristics match policy-defined high-scrutiny categories, engages all cognitive domains. That includes full empathy projection, integrity impact analysis across all three domains, and forecasting engine invocation with expanded branch depth. The thresholds and the high-scrutiny categories are policy-defined, so the boundary between shallow and deep engagement is set by governance rather than by the agent's momentary disposition.
Modulation by Other Cognitive Domains
The attention field is not static. It is modulated by a plurality of other cognitive domain fields through defined coupling pathways, so that the agent's allocation of evaluation effort responds to its own internal state and to its operating conditions. Elevated affective stress narrows the attention field toward immediate threats: engagement depth is reduced for domains not directly relevant to the stress-inducing condition, while engagement depth is increased for the confidence and capability domains that govern execution readiness under adverse conditions.
Recent integrity deviation broadens the attention field to include normative evaluation at maximum depth, ensuring that the agent's post-deviation mutations receive comprehensive integrity scrutiny rather than the abbreviated evaluation a low-stakes mutation would receive. A resource-constrained substrate, as reported by the capability envelope, narrows the attention field to reduce computational load, enabling the agent to maintain governance-compliant mutation evaluation within the available computational budget by deprioritizing non-critical evaluation dimensions. Detected operator distraction, as reported through the biological signal coupling mechanism, narrows the agent's attention field proportionally, reflecting the reduced oversight available from the human operator.
Attention as a Finite Resource
The premise that distinguishes the attention field is that evaluation is costly and bounded. A system that evaluated every proposed mutation across every cognitive domain at full depth would spend its computational budget indiscriminately, applying the same expensive scrutiny to a trivial mutation as to a consequential one. The attention field makes the allocation explicit and governable: it directs the deepest analyses, empathy projection, three-domain integrity impact analysis, and expanded-branch forecasting, toward the mutations whose projected impact justifies them, and it withholds those analyses from mutations that do not.
This is why the couplings to affective stress, integrity deviation, substrate resources, and operator distraction matter. Each coupling adjusts the allocation in response to a condition under which the default depth would be wrong: stress concentrates attention on execution readiness, a recent deviation forces full normative scrutiny, a constrained substrate sheds non-critical evaluation to stay within budget, and reduced human oversight narrows the field to match the reduced supervision available.
Tracking and Audit
The attention field is independently tracked, with a current value and a trajectory, and is recorded in the agent's lineage. Because the allocation for each mutation evaluation is recorded, the lineage supports post-hoc audit of which cognitive domains were consulted, and to what depth, for each mutation evaluation. An auditor reviewing a past mutation can therefore reconstruct not only the outcome of the evaluation but the depth of evaluation that produced it: whether a given mutation received full empathy projection and three-domain integrity analysis, or only the confidence and capability evaluation appropriate to a low-stakes change.
Tracking a trajectory in addition to a current value records how the allocation moved over time. Because the field is modulated by affective stress, integrity deviation, substrate resources, and operator distraction, its trajectory reflects the agent's changing conditions, and the lineage preserves the correspondence between those conditions and the engagement depth the agent applied while they obtained.
Composition With Mutation Evaluation
The attention field operates within the agent's mutation evaluation pipeline, the same pipeline in which the confidence governor, the integrity engine, and the capability envelope act as a composite admissibility evaluator over proposed mutations. The attention field does not replace that evaluation; it governs the depth at which the evaluating domains are consulted. By determining cognitive engagement depth per domain per cycle, it sits upstream of the analyses it allocates, deciding for each proposed mutation which of confidence, capability, empathy projection, integrity impact analysis, and forecasting will be invoked and how deeply.
Its couplings tie it to the other cognitive domain fields it allocates among and to the signals that should change that allocation. The affective state field narrows it under stress; the integrity field broadens it after a deviation; the capability envelope narrows it under resource constraint; and the biological signal coupling narrows it under operator distraction. Through these couplings the attention field keeps the cost of governance proportionate to the stakes of each mutation while remaining recorded and auditable in lineage.
Disclosure Scope
The attention field, as a cognitive domain field that governs which other cognitive domain fields are consulted and to what depth for a given mutation evaluation, representing cognitive engagement depth as a finite computational resource, is disclosed in the cognition filing (U.S. Application No. 19/647,395 and its international counterpart). The disclosure encompasses the determination of cognitive engagement depth as a scalar or vector quantity over the cognitive domain fields; the stakes-based differentiation in which low-impact mutations engage only confidence and capability evaluation while high-impact or high-scrutiny mutations engage all domains including full empathy projection, integrity impact analysis across all three domains, and forecasting engine invocation with expanded branch depth; the policy-defined low-impact and high-impact thresholds and high-scrutiny categories that set those boundaries; the modulation of the attention field by affective stress, recent integrity deviation, a resource-constrained substrate reported by the capability envelope, and detected operator distraction reported through biological signal coupling; and the independent tracking of the attention field's current value and trajectory in lineage to enable post-hoc audit of which cognitive domains were consulted and to what depth for each mutation evaluation. This article describes that disclosed mechanism and does not assert structures beyond it.
