Affective State as Seventh Canonical Field

Mechanism

A semantic agent in the foundational schema carries six canonical fields: an intent field, a context block, a memory field, a policy reference field, a mutation descriptor field, and a lineage field. These six fields let the agent carry its purpose, situational awareness, accumulated history, governance constraints, evolutionary record, and cryptographic provenance as a single persistent object. They do not, however, encode the agent's dispositional orientation toward its own execution environment: they do not capture the modulation state that shapes how the agent evaluates candidates, prioritizes mutations, or adjusts its responsiveness to environmental signals based on the cumulative outcomes of prior operations. The cognition disclosure adds the affective state field as a seventh structural field to close that gap.

The affective state field is a deterministic, policy-bounded data structure that encodes valence-weighted feedback derived from prior execution outcomes and environmental observations. It does not encode emotion in the phenomenological or subjective sense. It encodes a structured modulation vector that influences the agent's deliberation dynamics: how the agent weighs alternatives, tolerates ambiguity, persists under partial failure, and escalates under constraint pressure. The field is instantiated as one or more of a scalar valence value, a multi-dimensional vector of distinct modulation axes, or a structured record of named modulation fields each carrying magnitude, decay parameters, and policy bounds. In a fixed-schema embodiment, each named modulation field occupies a defined position and is independently readable, writable, and auditable.

Introducing affect as a seventh structural field, rather than as metadata, an annotation, or an external signal, is the load-bearing move. Because it is a field of the agent object, affective modulation participates in the same governance, lineage tracking, and policy enforcement that apply to every other field. Every mutation to the affective state field is recorded in the agent's lineage, subject to policy validation, and auditable by governance infrastructure. This structural integration distinguishes the disclosure from systems that treat affective state as a side channel, a prompt modifier, or an external behavioral overlay. The field is persisted with the agent across execution cycles, delegation events, and substrate migrations, so the agent's modulation state is not lost when the agent moves between execution environments or is serialized for transport.

The Named Control Fields

In one embodiment the affective state field is organized as a structured modulation layer comprising a plurality of named control fields. Each named control field encodes a distinct dimension of the agent's dispositional orientation and governs a specific aspect of deliberation behavior. The fields are not arbitrary labels: each corresponds to a measurable modulation axis with defined semantics, value ranges, update rules, and governance bounds. The disclosed control fields comprise at least the following.

An uncertainty sensitivity field, encoding responsiveness to epistemic uncertainty: elevated, the agent weights uncertain inputs more heavily and tends to defer or escalate; suppressed, it proceeds with greater tolerance for incomplete information. An ambiguity tolerance field, encoding capacity to operate where multiple interpretations remain unresolved: elevated, the agent maintains parallel candidate interpretations; reduced, it drives toward early disambiguation. A novelty appetite field, encoding disposition toward previously unobserved patterns, entities, or execution paths. A persistence-under-partial-failure field, encoding the tendency to continue a line of execution when intermediate results indicate partial failure or degraded confidence. An escalation-under-time-pressure field, encoding the tendency to escalate authority, request external input, or delegate under temporal constraints. A risk sensitivity field, encoding how the agent weights potential negative outcomes against positive ones. And a cooperation disposition field, encoding the tendency to favor collaborative strategies, including delegation, resource sharing, and multi-agent coordination, over independent execution.

In one embodiment each named control field is represented as a tuple: a current magnitude value within a defined range, a decay rate governing how rapidly the field returns toward a baseline in the absence of reinforcing stimuli, a policy-defined ceiling and floor bounding the permissible range, and a timestamp recording the most recent update.

What the Field Modulates

The affective state field modulates specific, enumerated targets within the agent's deliberation and execution pipeline. The targets are not open-ended; each is a defined computational parameter the field adjusts within governance bounds. The field does not create new capabilities, authorize new actions, or bypass policy constraints. It modulates the quantitative parameters that shape how existing authorized processes execute.

The disclosed targets include promotion thresholds, the minimum score required for a candidate to advance from one evaluation stage to the next, which elevated risk or uncertainty sensitivity raises; search breadth, the number of candidate alternatives explored at each decision point, which elevated novelty appetite or ambiguity tolerance widens; branch growth rates, the rate at which new speculative branches are generated during forecasting; decay rates for unpromoted candidates, the rate at which candidates not yet promoted are discarded from working memory, which elevated persistence slows; escalation thresholds, the conditions under which the agent transitions from independent operation to delegation or help-seeking; persistence parameters, the number of retry, reformulation, or alternative-strategy cycles undertaken before declaring failure; delegation routing preferences, the preference ordering among available delegates; and mutation acceptance thresholds, the stringency of the validation gate applied to mutations proposed by an external inference engine.

Deterministic, Policy-Bounded Updates

Updates to the affective state field are deterministic: given the same agent state, the same environmental inputs, and the same policy configuration, the update function produces the same output. This determinism is what makes the agent's affective evolution auditable, reproducible, and governable. In alternative embodiments the update function may incorporate bounded stochastic components, for example noise injection to simulate biological variance, provided the stochastic contribution is policy-bounded, recorded in lineage, and does not compromise the governance properties.

The update function operates on structured observations derived from the agent's execution environment. Disclosed observation categories include repeated failure patterns, which raise uncertainty and risk sensitivity; competing objectives, which raise ambiguity tolerance and escalation sensitivity; time pressure, which raises the escalation-under-time-pressure field and may narrow search breadth; novelty exposure, which raises or suppresses novelty appetite depending on whether prior novelty correlated with success or failure; uncertainty reported by an inference engine, which raises uncertainty sensitivity; and execution success patterns, which move the agent toward a more confident disposition. The function takes the current affective state vector, the current structured observations, and the applicable policy configuration, and produces an updated vector. Each dimension is updated independently according to its own rule, subject to policy bounds, and the update is recorded in lineage as a state mutation event with the input observations and resulting change preserved.

Every update is a policy-bounded mutation. The policy reference field specifies, per named control field, range bounds that clamp the resulting value to a floor and ceiling; rate limits that cap the magnitude of change per cycle so no single observation can produce a discontinuous jump; admissible triggers that define which observation types may drive a given field, so spurious or adversarial signals are ignored; update authority that defines which entities may initiate updates; and decay governance over the field's decay parameters. The update proceeds in stages: verify the observation type is admissible, compute the raw update, clamp to the rate limit, apply it, clamp the result to the range bounds, and record the complete transaction in lineage. This multi-stage clamping ensures no observation or sequence of observations can drive the field outside its policy-defined operating envelope.

Affect Does Not Override Governance

A strict separation of concerns is maintained between the affective modulation layer and the governance infrastructure. The affective state field cannot create authority the agent does not possess, bypass policy constraints, override trust slope validation, validate truth claims, or authorize execution that governance has denied. It modulates how the agent thinks, not whether the agent is permitted to act. The separation is enforced structurally: the governance gate evaluates execution admissibility based on policy compliance, trust slope validation, and cryptographic provenance independently of affective state, and the affective state field is not an input to that gate. Even with maximal confidence disposition and minimal risk sensitivity, the gate independently determines whether the action satisfies policy.

The separation operates across specific dimensions. On authority, an agent with elevated cooperation disposition still cannot delegate outside its policy-defined delegation scope, and one with elevated novelty appetite still cannot reach semantic neighborhoods its policy reference field excludes. On truth validation, suppressed uncertainty sensitivity does not cause uncertain information to be treated as verified; the agent's epistemic state is maintained independently of its dispositional state. On policy compliance, policy ceilings on field values and execution scope remain inviolable, enforced as hard constraints during every update. On trust slope validation, the cryptographic lineage trajectory is computed and validated independently of affect, and continuity requirements must be satisfied before execution regardless of affective configuration.

Decay, Hysteresis, and Quarantine

Each named control field is governed by a decay curve that returns the field toward its baseline in the absence of reinforcing stimuli, as a deterministic function of elapsed time, current deviation from baseline, and policy-specified decay parameters. In one embodiment decay is an exponential return toward a policy-defined resting value with a configurable time constant, and different fields may carry different time constants: uncertainty sensitivity may decay quickly because epistemic conditions change frequently, while persistence-under-partial-failure may decay slowly because it reflects deeper accumulated experience.

The modulation layer exhibits semantic hysteresis: current affective state depends on the trajectory of prior states, not only on current observations. Hysteresis is implemented through asymmetric update rules, where in one embodiment negative-valence updates driven by failure, uncertainty, or threat apply at a higher rate than positive-valence updates driven by success, producing a built-in caution bias in which the agent responds more rapidly to deteriorating conditions than it recovers from them. Entropy-governed valence stabilization prevents oscillatory behavior: when a field alternates rapidly between elevated and suppressed values, the stabilization mechanism progressively increases the effective decay time constant to damp the oscillation, with the stabilization threshold and damping factor set by policy.

Volatility that could compromise deliberation reliability routes the agent to an emotional quarantine state, a restricted execution mode rather than a suppression of affect. A volatility detector computes a composite metric from the recent update history of all named control fields; when it exceeds the quarantine threshold the agent enters quarantine, where promotion thresholds are raised to their maxima, delegation authority is suspended so the volatile state is not propagated to children, mutation acceptance thresholds are raised, and execution is routed through an additional validation layer. The agent continues to process observations and update its field while quarantined; it is released when the metric falls below a recovery threshold set below the quarantine threshold, the gap providing hysteresis that prevents oscillatory quarantine-release cycles. The mechanism is a circuit breaker for affective instability.

Composition Across the Cognitive Primitives

The affective state field is a cross-primitive input to confidence, forecasting, delegation, and discovery traversal. With respect to confidence, the field modulates the rate at which confidence decays and recovers: elevated uncertainty and risk sensitivity decay confidence faster, so the agent transitions sooner from executing mode to non-executing cognitive mode, while elevated persistence-under-partial-failure raises the recovery rate. With respect to forecasting, the engine reads the current affective state when initializing a planning graph generation cycle, with novelty appetite modulating branching factor, risk sensitivity modulating pruning, and persistence modulating projection depth. These dynamics are bounded by policy to prevent both runaway confidence and confidence collapse.

Affect composes with delegation through scoped inheritance. On a delegation event the parent's affective vector is evaluated against a policy-defined inheritance mask that marks each field inherited, excluded, or attenuated, and the inherited state is blended with the child's prior state under a policy-defined blending function, with the operation recorded in both lineages to establish a traceable affective provenance chain. Inheritance is depth-limited so modulation context does not propagate through arbitrarily deep chains, and a return mask may transmit part of the child's terminal state back to the parent. The same field also modulates discovery traversal: a discovery object's affective state shapes how it scores candidate transitions at each anchor, so an agent with elevated uncertainty sensitivity favors conservative transitions into familiar low-entropy neighborhoods while one with elevated novelty appetite favors transitions into uncharted territory, producing different but equally valid traversal paths through the same index for the same query.

Across multi-agent systems the field is governed by a formalized contagion model. Affective state propagates through delegation inheritance, interaction exposure during coordinated operations, and broadcast propagation from high-authority agents, with each channel processed through the standard policy-bounded update function. Anti-spiral mechanisms keep contagion bounded: a contagion damping factor strictly less than one attenuates influence at each propagation hop, per-field aggregate contagion limits cap the total shift within a time window, zone-level spiral detection activates additional suppression when a field's moving average deviates from baseline, and a quarantined agent's outgoing contagion is suppressed so it cannot propagate unstable state to the population.

Disclosure Scope

The affective state field as a seventh structural field of the semantic agent schema, comprising its deterministic policy-bounded data structure, the named control fields of the modulation layer (uncertainty sensitivity, ambiguity tolerance, novelty appetite, persistence-under-partial-failure, escalation-under-time-pressure, risk sensitivity, and cooperation disposition), the enumerated modulation targets, the deterministic update function over structured observations with multi-stage clamping and lineage recording, the separation of concerns that bars affect from creating authority or overriding governance, the decay curves, semantic hysteresis, entropy-governed stabilization, and emotional quarantine circuit breaker, and the cross-primitive composition with confidence, forecasting, scoped delegation inheritance, affect-modulated discovery traversal, and policy-bounded multi-agent contagion, is disclosed in the cognition filing (U.S. Application No. 19/647,395 and its international counterpart). This article describes that disclosed mechanism.

The scope extends to embodiments in which the field is realized as a scalar, a vector, or a named-field record; to deterministic update functions optionally carrying policy-bounded stochastic components recorded in lineage; to centralized, federated, decentralized, and embodied substrate realizations; and to biological-signal coupling and pseudonymous operation, provided in each case that updates remain deterministic, policy-bounded, lineage-recorded, and structurally separated from the governance gate.