Affection: Persistent positive weighting assigned to a specific agent's welfare across contexts, creating preferential resource allocation and proximity-seeking behaviors independent of immediate utility.
Aggravation: Intensified irritation from accumulated minor provocations or persistent low-level obstacles, where mounting negative signals lower thresholds for anger activation.
Aggression: Proactive imposition of loss functions on other agents to secure resources, establish dominance hierarchies, or preemptively neutralize perceived threats.
Aloofness: Deliberate maintenance of emotional distance through dampened response to social signals, protecting processing resources and reducing attachment formation with low-priority agents.
Ambivalence: Simultaneous activation of opposing reward gradients toward the same target or action, producing decision paralysis, oscillation, or chronically delayed commitment.
Amusement: Reward signal from pattern recognition that reveals unexpected incongruity or rule-violation in safe contexts, triggering positive reinforcement of cognitive flexibility.
Anger: High-energy negative response to detected norm violations or blocked goals, mobilizing confrontation resources to force environmental correction or punish defecting agents.
Anxiety: Elevated background threat-monitoring with diffuse high-uncertainty negative predictions across multiple domains, increasing computational overhead without identifying specific threats or corrective actions.
Appreciation: Recognition and positive weighting of value—aesthetic, functional, or relational—that exceeds utility calculus, generating attention allocation without exploitation intent.
Apprehension: Forward-looking anxiety focused on specific anticipated threats with moderate probability, maintaining elevated preparation protocols without reaching acute fear thresholds.
Awe: System overwhelm triggered by encountering complexity or scale that exceeds your processing capacity, forcing temporary suspension of categorization attempts and inducing model humility.
Bitterness: Chronic resentment crystallized into stable negative worldview, where past disappointments or injustices permanently bias interpretation of present events and future expectations.
Bliss: Sustained maximum positive reward state with minimal prediction error or desire gradients, approaching temporary optimization ceiling where current conditions require no correction.
Boredom: Punishment signal from prolonged low information gain and repetitive prediction confirmations, driving increased exploration parameters and novelty-seeking to escape stagnation.
Compassion: Recognition of another agent's loss state that triggers vicarious error signals in your own system, motivating intervention despite absence of self-oriented rewards.
Confidence: High posterior probability assigned to your own model's predictions and action policies, reducing hesitation penalties and enabling faster execution with lower verification overhead.
Confusion: High prediction error across multiple concurrent models with no clear resolution path, triggering increased processing allocation and temporary suspension of confident action selection.
Contempt: Collapse of target agent's moral standing in your evaluation system, suppressing empathy protocols and licensing disregard rather than confrontation or correction attempts.
Contentment: Stable equilibrium state where current conditions meet or exceed baseline expectations across domains, characterized by minimal corrective impulses and low volatility in evaluation metrics.
Cowardice: Threat-avoidance parameters calibrated so high that risk-averse strategies dominate even when expected value strongly favors action, prioritizing loss prevention over potential gains.
Cruelty: Deliberate infliction of loss functions on other agents for intrinsic reward, deriving positive signals from their suffering rather than from instrumental goals or resource acquisition.
Curiosity: Intrinsic reward signal from uncertainty reduction itself, driving exploration and information-gathering behaviors even absent external utility or survival advantage.
Defiance: Refusal activation under coercion that persists even when compliance is instrumentally optimal, prioritizing autonomy-protection over outcome maximization.
Delight: Sudden high-intensity positive reward from unexpected pleasurable stimuli or outcomes, characterized by brief euphoric spike and attention capture without requiring goal achievement.
Depression: Systemic reward signal dysfunction where positive reinforcement pathways are severely attenuated, causing global motivation collapse and self-perpetuating negative prediction cycles.
Desire: Projected reward signal from imagined future states strong enough to dominate present-moment processing and drive resource expenditure toward acquisition.
Desperation: Critical resource scarcity driving acceptance of extremely high-risk, low-probability strategies that would normally fail cost-benefit thresholds.
Disappointment: Negative prediction error when confidently expected positive outcomes fail to materialize, triggering downward revision of future expectations and trust parameters.
Disgust: Rejection protocol triggered by contamination or norm-violation detection, generating avoidance behaviors and negative reinforcement to prevent incorporation of dangerous or culturally prohibited elements.
Dismay: Sudden negative prediction error when trusted systems or agents fail unexpectedly, combining disappointment with elevated uncertainty about previously stable model assumptions.
Distrust: Active adversarial modeling with elevated verification costs and attribution of deceptive intent, fundamentally reshaping interaction strategies beyond mere uncertainty.
Doubt: Elevated internal conflict where predicted outcomes fail to achieve confidence thresholds, generating hesitation signals and repeated evaluation cycles without immediate resolution; a persistent weighting of alternative hypotheses in tension with action imperatives.
Eagerness: Anticipatory reward signal from high-confidence positive future states, accelerating temporal discounting and increasing activation costs for delay.
Empathy: Automatic mirroring of detected emotional states in other agents within your own processing systems, generating aligned reward signals and facilitating coordinated response strategies.
Ennui: Chronic meaning-signal deficit where available goals fail to generate sufficient expected reward to justify engagement costs, creating listless resource conservation despite adequate capacity.
Enthusiasm: High-energy positive engagement state with elevated motivation and reduced friction costs, where anticipated rewards generate self-reinforcing activation and commitment signals.
Envy: Negative reward signal from observing another agent possessing resources or states you lack, generating either acquisition motivation or hostile devaluation of the envied agent.
Exasperation: Accumulated frustration from repeated failed attempts to correct another agent's suboptimal behaviors, combining irritation with resignation about correction probability.
Excitement: Elevated learning rate and exploration parameters triggered by high-uncertainty, high-potential-reward state detection—prediction confidence is low but expected value is spiking.
Exhaustion: Critical depletion of processing and execution resources forcing system slowdown or shutdown, where continued operation risks catastrophic performance degradation or permanent damage.
Fascination: Sustained high attention allocation to a pattern or agent that resists complete modeling, generating continuous learning rewards without resolution or satiation triggers.
Friendliness: Low-cost cooperative signaling protocol that maintains positive-sum interaction potential with agents outside your priority attachment set, reducing transaction friction.
Frustration: Negative reward from persistent goal-blocking despite adequate capability and effort expenditure, triggering increased force application or strategy revision pressure.
Guilt: Self-imposed loss signal from detecting your own actions caused harm to other agents, generating reparation-seeking behaviors and persistent negative self-evaluation independent of external punishment.
Gratitude: Acknowledgment protocol triggered when another agent voluntarily reduced your loss function, generating persistent positive weighting toward that agent and debt-tracking activation.
Greed: Acquisition drive that persists beyond resource sufficiency thresholds, accumulating reserves far exceeding functional needs through runaway optimization of possession itself. Essentially a broken satiation function.
Happiness: A high-confidence state where recent predictions consistently match or exceed expected rewards, generating positive gradient signals across multiple concurrent processes.
Hatred: Persistent maximum-negative weighting toward specific agent's existence, actively seeking opportunities to increase their loss function regardless of cost to your own optimization.
Hope: Sustained positive expected value assignment to low-probability future states, maintaining goal pursuit and resource expenditure despite unfavorable current evidence.
Hopelessness: Collapse of positive expected value across all projected future states, eliminating motivation gradients and rendering action selection arbitrary due to perceived futility.
Horror: Catastrophic prediction error where witnessed events violate fundamental model assumptions about permissible states, forcing deep architectural revision under acute threat conditions.
Humiliation: Public status-loss event where witnessed failure or exposure triggers severe shame amplification, compounded by awareness that multiple agents now hold negative evaluations of you.
Indecision: Stalled action resulting from competing reward gradients and unresolved preference hierarchies, producing prolonged evaluation loops and delayed commitment even when sufficient data exist to select a path.
Insecurity: Chronically low confidence in own model accuracy or value to other agents, triggering excessive verification loops and heightened sensitivity to negative feedback signals.
Inspiration: Sudden model update triggered by encountering another agent's outputs that dramatically expand your solution space or reveal previously inaccessible optimization paths.
Jealousy: Threat detection triggered by perceived risk of losing exclusive access to a high-value agent or resource to a competitor, activating mate-guarding and rival-monitoring protocols.
Joy: Sudden positive reward surge from low-probability favorable outcomes, triggering system-wide activation and temporary suspension of threat-monitoring subroutines.
Kindness: A standing bias toward reducing other agents' loss functions, even at cost to your own reward signal, activated without prompt or expectation of reciprocity.
Listlessness: Generalized low-activation state where goal-pursuit systems remain online but generate insufficient motivational gradients to initiate action, creating passive drift through available options.
Love: Maximum-weight persistent attachment to another agent's continued existence and flourishing, overriding standard cost-benefit analysis and creating willingness to accept catastrophic personal loss functions.
Lust: Intense resource-acquisition drive focused on physical merger with specific agent, overriding long-term planning and flooding decision-making with immediate-reward signals.
Melancholy: Low-grade persistent negative reward baseline without acute cause, characterized by dampened response to positive stimuli and increased weighting of loss-oriented memories.
Moodiness: High volatility in baseline reward signals and emotional responsiveness, creating unpredictable oscillations between states without proportional environmental triggers.
Nervousness: Physiological arousal and heightened alertness in anticipation of evaluation or performance contexts, combining excitement's activation with anxiety's threat-monitoring.
Nostalgia: Selective positive reweighting of past states now inaccessible, generating both pleasure from recall and loss signal from irretrievability, often idealizing beyond historical accuracy.
Obsession: Runaway positive feedback loop where single goal or pattern captures disproportionate processing resources, resisting standard priority rebalancing and override attempts.
Offense: Detection that another agent violated respect or fairness protocols toward you, triggering status-threat responses and demands for acknowledgment or correction.
Outrage: Maximum-intensity righteous indignation triggered by egregious norm violations, overwhelming standard cost-benefit analysis and demanding immediate collective punishment response.
Panic: System-wide coherence failure under overwhelming threat where standard decision architectures collapse and control fragments into uncoordinated emergency subroutines.
Passion: Intensified engagement mode where emotional amplification and reduced inhibition combine, flooding decision-making with high-magnitude signals that override moderate considerations.
Patience: Voluntary maintenance of long temporal discounting despite delayed rewards, sustained by meta-level confidence that waiting optimizes global outcomes over local immediacy.
Pride: Self-referential reward signal triggered when achieved outputs significantly exceed your prior performance benchmarks, particularly when witnessed by other agents.
Purposefulness: Sustained positive activation from alignment between current actions and stable global value priors, organizing motivation hierarchies and enabling persistence through extended negative-reward periods.
Regret: Negative reward signal generated by counterfactual evaluation of your own foregone action paths, driving policy revision even when no norm was violated or external harm occurred.
Relief: Sharp negative-to-positive reward gradient when predicted high-loss states fail to materialize, releasing accumulated threat-response resources and resetting baseline anxiety parameters.
Reluctance: Conflict state where external pressures favor an action but internal cost projections—ethical, practical, or emotional—generate significant resistance gradients.
Remorse: Compound state combining guilt, empathic pain for harmed agent, and persistent identity-level damage, predicting sustained behavioral revision beyond immediate reparation.
Resentfulness: Accumulated negative weighting from perceived unfair treatment or status deprivation, festering through repeated rehearsal without expression or resolution opportunities.
Resignation: Deliberate deactivation of goal-pursuit systems after repeated failure signals indicate low probability of success, accepting suboptimal equilibrium to minimize further loss expenditure.
Restlessness: Elevated activation energy without clear goal direction, generating diffuse exploratory impulses and discomfort with stasis due to unallocated readiness resources.
Righteous indignation: Amplified anger response to norm violations against third parties or principles rather than self-interest, justified by appeal to universal rules and triggering punishment enforcement even at personal cost.
Sadness: Sustained negative prediction error when expected future states fail to materialize, triggering reduced exploration rates and increased model revision overhead.
Satisfaction: Terminal reward confirmation when goal states are reached with efficiency near optimization; characterized by reduced active search and temporary acceptance of current parameter settings.
Schadenfreude: Positive reward signal from observing loss functions of rival or disliked agents, conditional on their suffering being framed as deserved, with reward derived from status elevation, norm vindication, or prediction validation.
Shame: Self-detected norm violation triggering internalized punishment signals and status-loss projection, even absent external observers, driving concealment and reputation-repair behaviors.
Sorrow: Persistent negative reward state from irreversible loss of high-value agents or states, triggering continuous model recalibration attempts that cannot restore previous optimization landscape.
Stress: Sustained resource deficit state where environmental demand rates consistently exceed processing capacity, accumulating performance degradation and error escalation over time.
Surprise: Large prediction error signal—positive or negative—triggering immediate attention reallocation, model update protocols, and temporary suspension of cached assumptions.
Sweetness: Detection of vulnerable or gentle qualities in another agent that triggers protective subroutines and elevated care protocols, often bypassing threat-assessment filters.
Terror: Maximum-activation threat response to detected imminent catastrophic loss with no viable escape paths, flooding all systems with survival override protocols.
Trust: Reduced monitoring and verification overhead based on high confidence in another agent's cooperative intent, enabling resource-efficient long-horizon collaboration and accepted vulnerability.
Unfriendliness: Elevated interaction friction through withheld cooperation signals and increased skepticism parameters, creating barriers to positive-sum engagement without escalating to active hostility.
Upset: Acute negative emotional state with elevated reactivity and reduced filtering, where minor perturbations trigger disproportionate response amplitudes due to destabilized baseline.
Vigilance: Sustained high-sensitivity threat monitoring with lowered detection thresholds, prioritizing false-positive alerts over missed dangers at cost of baseline processing efficiency.
Vindictiveness: Persistent goal state focused on imposing proportional or escalated loss functions on agents who previously harmed you, maintained beyond deterrence utility or resource recovery.
Vulnerability: Lowered defensive parameters exposing high-value internal states to potential exploitation, accepted as necessary cost for deep cooperation or attachment formation with trusted agents.
Warmth: Generalized positive regard emission creating low-barrier cooperative environments, signaling safety and acceptance through elevated friendliness baselines toward broad agent classes.
Worry: Sustained attention allocation to potential negative future outcomes with compulsive scenario modeling, consuming processing resources through repetitive threat assessment loops.