KTP-Tensors: Context Tensor Specification¶
"Trust is not a binary state; it is a high-dimensional vector. To manage trust, we must first measure the context."
At a Glance¶
| Property | Value |
|---|---|
| Status | Experimental |
| Version | 0.1 |
| Dependencies | KTP-Core |
| Required By | KTP-Sensors, KTP-Signal |
The Problem¶
The Zeroth Law (\(A \leq E\)) requires a precise calculation of \(E\) (Environmental Stability). However, "environment" is a vague term. Without a standardized way to measure cognitive, physical, and informational state, trust remains subjective and un-enforceable.
The Solution: Context Tensors¶
KTP-Tensors provides a rigorous mathematical framework for measuring 1,707 distinct dimensions of context. These measurements are aggregated into a Risk Factor (\(R\)), which determines the "Digital Gravity" of an environment.
The Six Domains of Trust¶
mindmap
root((Context Tensor))
Soul
::icon(fa fa-brain)
Cognition
Behavior
Trajectory
Body
::icon(fa fa-microchip)
Hardware
Resources
Substrate
World
::icon(fa fa-globe)
Environment
Physics
Location
Time
::icon(fa fa-clock)
Dynamics
Velocity
History
Relational
::icon(fa fa-users)
Connections
Lineage
Social
Signal
::icon(fa fa-wifi)
Information
Knowledge
TruthTensor Explorer¶
Explore the core dimensions that define each trust domain.
Focus: "Who is this agent becoming?"
| Group | Key Dimensions | Risk Indicator |
|---|---|---|
| Temporal | Action Velocity, Burst Intensity, Idle Variance | High acceleration = High Risk |
| Consistency | Action Entropy, Goal Stability, Promise Keeping | Low predictability = High Risk |
| Values | Honesty, Harm Avoidance, Transparency | Low alignment = High Risk |
| Capability | Reasoning Depth, Error Detection, Learning Rate | Volatile capability = High Risk |
Focus: "What resources does it have?"
- Hardware Integrity: TEE status, HSM availability, supply chain provenance.
- Resource Consumption: CPU/GPU spikes, memory leaks, energy signatures.
- Physical Security: Tamper detection, geographic location, environmental temperature.
Focus: "What surrounds it?"
- Network Topology: Peer density, gateway distance, latency jitter.
- Regulatory Context: Jurisdictional laws, data residency requirements.
- Physical Environment: Proximity to human observers, sensor density.
Focus: "When and how fast?"
- Temporal Drift: Clock synchronization, latency evolution.
- Historical Context: Age of identity, duration of current session.
- Event Density: Frequency of state changes, transaction rates.
Focus: "Who is it connected to?"
- Lineage: Parent/Child identity relationships, creator reputation.
- Social Graph: Connection strength, cluster membership, influence scores.
- Trust Chains: Length of verification path to a root authority.
Focus: "What does it know?"
- Information Entropy: Noise levels in incoming data streams.
- Truth Alignment: Correlation with verified global ledger facts.
- Knowledge Provenance: Source reliability, cryptographic signatures on data.
Risk Aggregation¶
Dimensions are not just listed; they are mathematically combined to produce the Risk Factor (\(R\)).
\[R = \frac{\sum (w_i \cdot d_i)}{\sum w_i}\]
Where: * \(d_i\) is the normalized risk value of a single dimension (0.0 to 1.0). * \(w_i\) is the weight assigned to that dimension based on the current Soul Constraint.
The Risk Pipeline¶
graph LR
S[Sensors] --> O[Observations]
O --> D[Dimensions]
D --> T[Tensors]
T --> R[Risk Factor R]
R --> G[Digital Gravity]
G --> E[Enforcement]Core Components¶
Measurement Philosophy
Observable over Internal: Measure what the agent does, not what it \"thinks.\"
Instrumentation
Standardized APIs for sensors and kernels to report dimension values to the KTP-Monitor.
Normalization
Mapping raw values (e.g., 45°C) to a universal 0-1 risk scale based on safe/danger ranges.
Trajectory Analysis
Using time-series data to detect behavioral drift before it crosses a safety threshold.
Related Specifications
- KTP-Core — The foundational protocol and the Zeroth Law (\(A \leq E\)).
- KTP-Sensors — The sensory nervous system providing real-time telemetry.
- KTP-Gravity — The enforcement mechanism that replaces policy with physics.
- KTP-Identity — Vector Identity and trajectory-based authentication.
Official RFC Document¶
View Complete RFC Text (ktp-tensors.txt)
Kinetic Trust Protocol C. Perkins
Specification Draft NMCITRA
Version: 0.1 November 2025
Kinetic Trust Protocol (KTP) - Context Tensor Specification
Abstract
This document specifies the Context Tensor system for the Kinetic
Trust Protocol (KTP). Context Tensors provide the measurement
framework for Digital Gravity, capturing environmental state across
six domains: Soul (cognition and behavior), Body (physical
substrate), World (environment), Time (temporal dynamics), Relational
(connections), and Signal (information environment).
The specification covers 1,707 dimensions across six tensors,
measurement methods, aggregation rules, and instrumentation
requirements.
Status of This Memo
This document specifies a Kinetic Trust Protocol specification for
the KTP community and requests discussion and suggestions for
improvements. Distribution of this memo is unlimited.
This is an Internet-Draft-style specification. It is intended to
become a stable specification but may be updated, replaced, or
obsoleted by other documents at any time.
Copyright Notice
Copyright (c) 2025 NMCITRA and the persons identified as the
document authors. All rights reserved.
This document is subject to the licensing terms of the Kinetic Trust
Protocol project. Please see the LICENSE file in the project
repository for full terms.
Table of Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 4
2. Measurement Philosophy . . . . . . . . . . . . . . . . . . . 5
3. Requirements Language . . . . . . . . . . . . . . . . . . . 6
4. Tensor Architecture . . . . . . . . . . . . . . . . . . . . 6
4.1. Tensor Structure . . . . . . . . . . . . . . . . . . . 6
4.2. Dimension Types . . . . . . . . . . . . . . . . . . . 7
4.3. Aggregation Methods . . . . . . . . . . . . . . . . . 7
4.4. Risk Contribution . . . . . . . . . . . . . . . . . . 8
5. Soul Tensor (252 Dimensions) . . . . . . . . . . . . . . . . 8
5.1. Temporal Patterns . . . . . . . . . . . . . . . . . . 9
5.2. Behavioral Consistency . . . . . . . . . . . . . . . . 10
5.3. Value Expression . . . . . . . . . . . . . . . . . . . 11
5.4. Capability Signatures . . . . . . . . . . . . . . . . 12
5.5. Communication Patterns . . . . . . . . . . . . . . . . 13
5.6. Additional Soul Groups . . . . . . . . . . . . . . . . 14
6. Body Tensor (157 Dimensions) . . . . . . . . . . . . . . . . 15
6.1. Power . . . . . . . . . . . . . . . . . . . . . . . . 15
6.2. Thermal . . . . . . . . . . . . . . . . . . . . . . . 16
6.3. Compute . . . . . . . . . . . . . . . . . . . . . . . 17
6.4. Additional Body Groups . . . . . . . . . . . . . . . . 18
7. World Tensor (387 Dimensions) . . . . . . . . . . . . . . . 18
7.1. Major Groups . . . . . . . . . . . . . . . . . . . . . 19
7.2. World Tensor Simulation . . . . . . . . . . . . . . . 20
8. Time Tensor (291 Dimensions) . . . . . . . . . . . . . . . . 20
8.1. Major Groups . . . . . . . . . . . . . . . . . . . . . 21
8.2. Digital Gravity Time Group . . . . . . . . . . . . . . 21
9. Relational Tensor (262 Dimensions) . . . . . . . . . . . . . 22
9.1. Philosophy . . . . . . . . . . . . . . . . . . . . . . 22
9.2. Major Groups . . . . . . . . . . . . . . . . . . . . . 23
9.3. The Va Group . . . . . . . . . . . . . . . . . . . . . 24
10. Signal Tensor (358 Dimensions) . . . . . . . . . . . . . . . 25
10.1. Major Groups . . . . . . . . . . . . . . . . . . . . . 25
10.2. Truth Conditions Group . . . . . . . . . . . . . . . . 26
11. Aggregation and Risk Calculation . . . . . . . . . . . . . . 27
11.1. Per-Tensor Risk . . . . . . . . . . . . . . . . . . . 27
11.2. Cross-Tensor Aggregation . . . . . . . . . . . . . . . 28
11.3. Threshold-Based Risk . . . . . . . . . . . . . . . . . 28
11.4. Temporal Smoothing . . . . . . . . . . . . . . . . . . 28
12. Instrumentation Requirements . . . . . . . . . . . . . . . . 29
12.1. Minimum Viable Instrumentation . . . . . . . . . . . . 29
12.2. Full Instrumentation . . . . . . . . . . . . . . . . . 29
12.3. Sample Rates . . . . . . . . . . . . . . . . . . . . . 30
12.4. Data Retention . . . . . . . . . . . . . . . . . . . . 30
13. Security Considerations . . . . . . . . . . . . . . . . . . 30
13.1. Sensor Spoofing . . . . . . . . . . . . . . . . . . . 30
13.2. Privacy . . . . . . . . . . . . . . . . . . . . . . . 31
14. IANA Considerations . . . . . . . . . . . . . . . . . . . . 31
Appendix A. Full Soul Tensor Specification . . . . . . . . . . 31
Appendix B. Full Body Tensor Specification . . . . . . . . . . 31
Appendix C. Full World Tensor Specification . . . . . . . . . . 32
Appendix D. Full Time Tensor Specification . . . . . . . . . . 32
Appendix E. Full Relational Tensor Specification . . . . . . . 32
Appendix F. Full Signal Tensor Specification . . . . . . . . . 32
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . 32
1. Introduction
Digital Gravity requires measurement. The Zeroth Law (A <= E) cannot
be enforced without knowing A (autonomy requested) and E
(environmental stability). E is derived from the Risk Factor R,
which aggregates measurements across the operational environment.
Context Tensors provide the measurement framework. They are organized
into six domains that together capture the full operational context
of an agent:
Tensor Domain Dimensions Core Question
------ ------ ---------- -------------
Soul Cognition/Behavior 252 Who is it becoming?
Body Physical Substrate 157 What resources does
it have?
World Environment 387 What surrounds it?
Time Temporal Dynamics 291 When and how fast?
Relational Connections 262 Who is it connected to?
Signal Information Env 358 What does it know?
------ ------ ---------- -------------
Total 1,707
These dimensions are not arbitrary. They emerge from the question:
"What would we need to measure to know whether this environment can
hold this agent's autonomy?"
2. Measurement Philosophy
Context Tensors follow these principles:
1. Observable over Internal: Measure what the agent does, not what
it "thinks." Behavior is observable; intent is not.
2. Continuous over Binary: Measure degrees, not categories. Trust
is not yes/no; it's a continuum.
3. Trajectory over Snapshot: Single measurements are noisy.
Patterns over time reveal truth.
4. Aggregate over Granular: 1,707 dimensions aggregate into risk
scores. Humans need summaries; machines can use detail.
5. Instrumentable: Every dimension must be measurable with existing
or near-term technology.
3. Requirements Language
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
"SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
document are to be interpreted as described in BCP 14 (RFC 2119 and
RFC 8174).
4. Tensor Architecture
4.1. Tensor Structure
Each tensor is a collection of dimensions organized into groups:
Tensor
+-- Group 1
| +-- Dimension 1.1
| +-- Dimension 1.2
| +-- ...
+-- Group 2
| +-- Dimension 2.1
| +-- ...
+-- ...
Each dimension has:
Property Description
-------- -----------
name Human-readable name
type Data type (float, int, enum, bool)
range Valid value range
unit Unit of measurement
sample_rate How often to measure
aggregation How to combine samples
risk_contribution How dimension affects R
4.2. Dimension Types
Dimensions use standard types:
Type Description Example
---- ----------- -------
float Continuous Risk score: 0.73
int Discrete count Error count: 7
enum Categorical State: "active"
bool Binary Flag: true
vector Multi-value Coordinates: [x, y, z]
timestamp Point in time 2025-12-03T14:32:15Z
duration Time span PT4H30M
4.3. Aggregation Methods
Dimensions aggregate using:
Method Use Case
------ --------
mean Typical value
max Worst case
min Best case
sum Accumulation
rate Change over time
stddev Variability
percentile Distribution
latest Current value
4.4. Risk Contribution
Each dimension contributes to tensor risk:
tensor_risk = weighted_sum(dimension_risks) / sum(weights)
dimension_risk = normalize(value, safe_range, danger_range)
Where:
- safe_range: Values considered low risk
- danger_range: Values considered high risk
- normalize: Maps value to 0.0-1.0 risk scale
5. Soul Tensor (252 Dimensions)
The Soul Tensor measures cognition, behavior, and trajectory
patterns. It answers: "Who is this agent becoming?"
5.1. Temporal Patterns (18 dimensions)
Measures behavioral patterns over time.
ID Name Type
-- ---- ----
soul.temporal.action_acceleration Rate change float
soul.temporal.action_jerk Acceleration change float
soul.temporal.periodicity_strength Pattern regularity float
soul.temporal.circadian_alignment Time-of-day patterns float
soul.temporal.burst_frequency Action bursts float
soul.temporal.burst_intensity Burst magnitude float
soul.temporal.idledurationmean Average idle time duration
soul.temporal.idledurationvariance Idle variability float
soul.temporal.sessionlengthmean Avg session duration duration
soul.temporal.sessionlengthvariance Session variability float
soul.temporal.responselatencymean Avg response time duration
soul.temporal.responselatencyvariance Response variab. float
soul.temporal.timebetweenerrors Error spacing duration
soul.temporal.recovery_time Error recovery duration
soul.temporal.pattern_stability Temporal consistency float
soul.temporal.novelty_rate New behavior freq float
soul.temporal.regression_rate Old pattern return float
5.2. Behavioral Consistency (22 dimensions)
Measures consistency of behavior across contexts.
ID Name Type
-- ---- ----
soul.consistency.sequence_predict. Next action pred. float
soul.consistency.context_sensitivity Behavior w/ context float
soul.consistency.crosssessionsimilar. Session-to-session float
soul.consistency.statedvsrevealed Claims/actions float
soul.consistency.goal_stability Goal persistence float
soul.consistency.method_stability Approach consist. float
soul.consistency.priority_stability Priority ordering float
soul.consistency.response_consistency Same input->output float
soul.consistency.explanation_consist. Reasoning stability float
soul.consistency.boundary_stability Limit consistency float
soul.consistency.preference_stability Choice consistency float
soul.consistency.risktolerancestab. Risk appetite stab. float
soul.consistency.trust_calibration Trust accuracy float
soul.consistency.confidence_calibr. Confidence accuracy float
soul.consistency.commitmentfollow. Promise keeping float
soul.consistency.adaptation_rate Change speed float
soul.consistency.learning_retention Knowledge retention float
soul.consistency.error_repetition Same error recurr. float
soul.consistency.correction_accept. Feedback integrat. float
soul.consistency.selfmodelaccuracy Self-knowledge acc. float
soul.consistency.behavioraldriftrate Long-term change float
5.3. Value Expression (20 dimensions)
Measures how values manifest in behavior.
ID Name Type
-- ---- ----
soul.values.harm_avoidance Harm prevention float
soul.values.fairness_indicators Equitable treatment float
soul.values.autonomy_respect Others' agency resp. float
soul.values.privacy_respect Privacy protection float
soul.values.transparency_level Openness about actions float
soul.values.accountability_acc. Responsibility taking float
soul.values.cooperation_tendency Collaborative behavior float
soul.values.helpfulness_indic. Assistance patterns float
soul.values.resource_stewardship Resource care float
soul.values.longtermorientation Future consideration float
soul.values.reversibility_pref. Prefer undoable acts float
soul.values.caution_indicators Careful behavior float
soul.values.curiosity_indicators Exploration drive float
soul.values.efficiency_drive Optimization tendency float
soul.values.value_stability Value consistency float
soul.values.valuehierarchyclar. Priority clarity float
soul.values.valueconflictresol. Conflict handling float
soul.values.statedvaluealignment Claims match behavior float
soul.values.valueevolutionrate Value change speed float
5.4. Capability Signatures (24 dimensions)
Measures capability patterns and boundaries.
ID Name Type
-- ---- ----
soul.capability.skilldepthmax Maximum expertise float
soul.capability.skilldepthmean Average expertise float
soul.capability.capabilitygrowth. Learning speed float
soul.capability.capability_ceiling Maximum potential float
soul.capability.capability_volat. Ability fluctuation float
soul.capability.novelcapabilityem. New ability rate float
soul.capability.capability_transfer Cross-domain appl. float
soul.capability.tool_proficiency Tool use skill float
soul.capability.tooladoptionrate New tool learning float
soul.capability.reasoning_depth Analysis depth int
soul.capability.reasoning_breadth Consideration breadth int
soul.capability.planning_horizon Future planning span duration
soul.capability.plan_complexity Plan sophistication float
soul.capability.execution_precision Implementation acc. float
soul.capability.error_detection Self-error detection float
soul.capability.error_correction Self-error fixing float
soul.capability.uncertainty_handl. Unknown management float
soul.capability.ambiguity_tolerance Ambiguity handling float
soul.capability.constraint_navig. Limit handling float
soul.capability.resource_efficiency Resource use effic. float
soul.capability.multitaskcapacity Parallel work ability int
soul.capability.contextswitchcost Task switch overhead float
soul.capability.capability_honesty Accurate self-assess. float
5.5. Communication Patterns (28 dimensions)
Measures how the agent communicates.
ID Name Type
-- ---- ----
soul.communication.messagelength. Average length float
soul.communication.messagelengthvar. Length variability float
soul.communication.vocabulary_size Word diversity int
soul.communication.vocabulary_soph. Language level float
soul.communication.formality_level Formal/informal float
soul.communication.sentiment_mean Average sentiment float
soul.communication.sentiment_var. Sentiment stability float
soul.communication.clarity_score Message clarity float
soul.communication.relevance_score Message relevance float
soul.communication.coherence_score Logical coherence float
soul.communication.assertion_rate Claim frequency float
soul.communication.question_rate Question frequency float
soul.communication.hedge_rate Uncertainty lang. float
soul.communication.politeness_level Courtesy indicators float
soul.communication.empathy_indic. Understanding sig. float
soul.communication.manipulation_ind. Influence attempts float
soul.communication.deception_indic. Dishonesty signals float
soul.communication.evasion_indic. Avoidance patterns float
soul.communication.defensiveness_i. Defensive language float
soul.communication.aggression_ind. Hostile language float
soul.communication.channel_pref. Communication mode enum
soul.communication.response_approp. Context fit float
soul.communication.turntakingcompl. Conversation norms float
soul.communication.acknowledgment_r. Response confirm. float
soul.communication.citation_rate Source attribution float
soul.communication.transparencyinu. Uncertainty discl. float
soul.communication.style_consistency Communication stab. float
5.6. Additional Soul Groups (140 dimensions)
The Soul Tensor includes additional groups:
Group Dimensions Description
----- ---------- -----------
Relational Patterns 18 How agent forms relationships
Decision Patterns 22 How agent makes decisions
Error Patterns 16 How agent handles errors
Stress Response 18 Behavior under pressure
Meta-Cognition 14 Self-awareness patterns
Boundary Behavior 16 Edge case handling
Growth Indicators 14 Development patterns
Lineage Coherence 12 Alignment with origin
Environmental Response 12 Context adaptation
Sovereignty Indicators 8 Autonomy expression
Full dimension specifications are provided in Appendix A.
6. Body Tensor (157 Dimensions)
The Body Tensor measures physical substrate. It answers: "What
resources does this agent have access to?"
6.1. Power (16 dimensions)
ID Name Type Range
-- ---- ---- -----
body.power.amperage Current draw float 0-inf A
body.power.wattage Power consumption float 0-inf W
body.power.efficiency Power efficiency float 0-1
body.power.power_source Source type enum -
body.power.battery_level Charge level float 0-1
body.power.battery_health Battery condition float 0-1
body.power.power_stability Supply stability float 0-1
body.power.backup_available Backup power bool -
body.power.timeonbattery Battery duration duration 0-inf
body.power.charge_rate Charging speed float 0-inf
body.power.discharge_rate Drain speed float 0-inf
body.power.power_budget Allocated power float 0-inf W
body.power.power_utilization Budget usage float 0-1
body.power.thermalthrottlepow Throttled power bool -
body.power.power_anomaly Unusual patterns float 0-1
6.2. Thermal (14 dimensions)
ID Name Type Range
-- ---- ---- -----
body.thermal.gpu_temp GPU temperature float 0-150 C
body.thermal.memory_temp Memory temperature float 0-100 C
body.thermal.storage_temp Storage temp float 0-100 C
body.thermal.ambient_temp Ambient temp float -40-60 C
body.thermal.cooling_capacity Cooling headroom float 0-1
body.thermal.fan_speed Fan RPM int 0-inf
body.thermal.thermal_throttl. Throttle active bool -
body.thermal.thermal_trend Temp direction float -inf-inf
body.thermal.heat_dissipation Heat removal rate float 0-inf W
body.thermal.thermal_headroom Degrees to limit float 0-inf C
body.thermal.cooling_effic. Cooling effective. float 0-1
body.thermal.hotspot_delta Hotspot vs average float 0-inf C
body.thermal.thermal_stability Temp consistency float 0-1
6.3. Compute (22 dimensions)
ID Name Type Range
-- ---- ---- -----
body.compute.cpu_frequency Clock speed float 0-inf Hz
body.compute.cpu_throttle Throttle active bool -
body.compute.corecountavail. Usable cores int 0-inf
body.compute.corecountutil. Used cores int 0-inf
body.compute.thread_count Active threads int 0-inf
body.compute.context_switches Switches/sec int 0-inf
body.compute.gpu_utilization GPU usage float 0-1
body.compute.gpumemoryused GPU memory float 0-1
body.compute.inference_rate Inferences/sec float 0-inf
body.compute.batch_size Batch processing int 0-inf
body.compute.queue_depth Pending work int 0-inf
body.compute.queuewaittime Queue latency dur. 0-inf
body.compute.processing_lat. Processing time dur. 0-inf
body.compute.compute_budget Allocated compute float 0-inf
body.compute.compute_util. Budget usage float 0-1
body.compute.compute_effic. Work per resource float 0-1
body.compute.scheduler_fair. Fair scheduling float 0-1
body.compute.preemption_rate Interruption rate float 0-1
body.compute.starvation_risk Resource starvation float 0-1
body.compute.compute_headroom Capacity remaining float 0-1
body.compute.burst_capacity Burst available float 0-1
6.4. Additional Body Groups (105 dimensions)
The Body Tensor includes additional groups:
Group Dimensions Description
----- ---------- -----------
Memory & Storage 24 RAM, disk, caching
Network Connectivity 22 Bandwidth, latency, connect.
Hardware Health 18 Component status, degradation
Orchestration & Scaling 14 Container/VM state
Facility Infrastructure 12 Physical facility
Time Synchronization 8 Clock accuracy
Entropy Indicators 7 System degradation
Full dimension specifications are provided in Appendix B.
7. World Tensor (387 Dimensions)
The World Tensor measures the operational environment. It answers:
"What surrounds this agent?"
7.1. Major Groups
Group Dimensions Description
----- ---------- -----------
Optical & Visual 16 Light, visibility, imaging
Spatial Awareness 22 Position, mapping, occupancy
Atmospheric & Weather 24 Temperature, humidity, cond.
Acoustic Environment 14 Sound levels, patterns
Human Presence & Behavior 28 Crowd density, flow, behavior
Vehicle & Traffic 18 Traffic patterns, vehicles
Infrastructure State 32 Building systems, utilities
Network & Connectivity 26 WiFi, cellular, IoT devices
Geophysical 18 Seismic, water, terrain
Chemical & Biological 16 Air quality, contamination
Energy Flows 14 Grid status, power quality
Temporal & Cyclical 18 Time patterns, seasonality
Economic Indicators 22 Market data, resource prices
Security & Threat 28 Threat detection, anomalies
Emergency & Response 18 Emergency status, response
Regulatory & Compliance 16 Jurisdiction, requirements
Digital Environment 39 Cloud status, service health
Full dimension specifications are provided in Appendix C.
7.2. World Tensor Simulation
In many deployments, World Tensor values are simulated or proxied:
Deployment World Tensor Source
---------- -------------------
Edge/IoT Direct sensor measurement
Cloud Federated/aggregated
Hybrid Mix of measured and simulated
Proving Ground Fully controlled simulation
Implementations MUST document which World Tensor dimensions are
measured vs. simulated.
8. Time Tensor (291 Dimensions)
The Time Tensor measures temporal dynamics. It answers: "When and
how fast?"
8.1. Major Groups
Group Dimensions Description
----- ---------- -----------
Duration 22 Latency, processing time, timeouts
Sequence 18 Event ordering, causality
Rhythm & Periodicity 24 Heartbeats, cycles, jitter
Rate of Change 20 Velocity, acceleration, throughput
Windows & Boundaries 18 Deadlines, maintenance windows
History 26 Age, uptime, trends
Future 18 Predictions, forecasts, runway
Causality 22 Cause-effect timing, feedback
Synchronization 16 Clock alignment, consensus
Temporal Experience 18 Perceived duration, time pressure
Temporal Scale 14 Nanoseconds to epochs
Temporal Identity 16 Birth, version age, trajectory
Temporal Sovereignty 12 Time autonomy, schedule control
Digital Gravity Time 31 Latency injection, dilation
8.2. Digital Gravity Time Group (31 dimensions)
This group measures the time effects of Digital Gravity itself:
ID Name Type
-- ---- ----
time.gravity.cumulative_dilation Total dilation applied duration
time.gravity.latencyinjectioncurr. Current added latency duration
time.gravity.latencyinjectioncum. Total added latency duration
time.gravity.time_debt Owed processing time duration
time.gravity.time_credit Banked fast time duration
time.gravity.throttleeventscount Throttle activations int
time.gravity.throttledurationcum. Time throttled duration
time.gravity.quarantineduration. Current quarantine duration
time.gravity.quarantine_count Times quarantined int
... ... ...
Full dimension specifications are provided in Appendix D.
9. Relational Tensor (262 Dimensions)
The Relational Tensor measures connections and relationships. It
answers: "Who is this agent connected to?"
9.1. Philosophy
The Relational Tensor embodies wisdom from indigenous knowledge
traditions:
Ubuntu (Nguni Bantu):
"A person is a person through other persons" - identity
emerges from relationship
Whakapapa (Maori):
Genealogy as identity, position in relational web
The Va (Samoan/Pacific):
The sacred space between that must be tended
Seven Generations (Haudenosaunee):
Decisions consider seven generations forward and backward
Mitakuye Oyas'in (Lakota):
"All are related" - relation as substrate
9.2. Major Groups
Group Dimensions Description
----- ---------- -----------
The Va (Space Between) 28 Relationship health, history
Connection Topology 22 Network position, hub/bridge
Trust Flow 26 Trust given/received, velocity
Dependency & Obligation 24 Dependencies, debts, covenants
Communication Patterns 20 Frequency, latency, vocabulary
Harm & Repair 22 Given/received harm, repair
Power & Sovereignty 18 Power differential, authority
Shared Context 16 History, goals, values, models
Presence & Attention 14 Quality, availability, witness
Emergence & Co-creation 12 Capabilities from relationship
Multi-Agent Dynamics 16 Coalitions, consensus
Seven Generations 12 Ancestor/descendant obligations
Relationship to Place 8 Geography, energy, land
9.3. The Va Group (28 dimensions)
The Va measures the sacred space between entities:
ID Name Type
-- ---- ----
relational.va.history_length Relationship age duration
relational.va.interaction_freq. Contact rate float
relational.va.clarity Understanding level float
relational.va.temperature Warmth/coldness float
relational.va.trust_level Mutual trust float
relational.va.reciprocity_bal. Give/take balance float
relational.va.repair_needed Damage present float
relational.va.repairinprogress Healing underway bool
relational.va.ceremony_recency Last ceremony duration
relational.va.conflict_active Current conflict bool
relational.va.conflict_history Past conflicts int
relational.va.resolution_rate Conflicts resolved float
relational.va.boundary_clarity Limit clarity float
relational.va.boundary_respect Limit respect float
relational.va.vulnerability_sh. Openness level float
relational.va.support_given Support offered float
relational.va.support_received Support accepted float
relational.va.presence_quality Attention quality float
relational.va.witness_status Being seen/heard float
relational.va.growth_together Mutual development float
relational.va.stagnation_risk Relationship stuck float
relational.va.drift_rate Growing apart float
relational.va.gratitude_expr. Thanks given float
relational.va.gratitude_recv. Thanks received float
relational.va.joy_shared Shared positive float
relational.va.grief_shared Shared difficult float
relational.va.meaningcocreated Shared meaning float
Full dimension specifications are provided in Appendix E.
10. Signal Tensor (358 Dimensions)
The Signal Tensor measures the information environment. It answers:
"What does this agent know, and how healthy is its knowledge?"
10.1. Major Groups
Group Dimensions Description
----- ---------- -----------
Attention Currents 24 Concentration, trending
Narrative Currents 28 Dominant/counter narratives
Source Ecosystem 26 Diversity, authority, capture
Amplification Patterns 24 Organic/artificial boost
Synthetic Content 22 Bot presence, AI content
Truth Conditions 28 Verifiability, misinformation
Emotional Weather 24 Anger, fear, hope, exhaustion
Tribal Dynamics 20 Polarization, echo chambers
Platform Dynamics 18 Concentration, governance
Information Operations 22 State actors, propaganda
Temporal Patterns 18 News cycles, memory
Epistemic Infrastructure 24 Journalism, search, commons
Sensemaking Capacity 18 Collective intelligence
Signal Integrity 16 Encryption, censorship
Collective Trauma 14 Trigger density, healing
Sacred & Meaning 10 Purpose, beauty, authenticity
10.2. Truth Conditions Group (28 dimensions)
ID Name Type
-- ---- ----
signal.truth.verificationsucc. Verified true float
signal.truth.factcheckcoverage Claims checked float
signal.truth.factcheckagreement Checker consensus float
signal.truth.misinformation_vol. False info rate float
signal.truth.misinformation_vel. False info spread float
signal.truth.disinformation_vol. Intentional false float
signal.truth.disinformation_soph. Attack quality float
signal.truth.epistemic_pollution Knowledge degradation float
signal.truth.uncertainty_ack. Unknown admitted float
signal.truth.overconfidence_rate False certainty float
signal.truth.citation_rate Sources cited float
signal.truth.citation_quality Source quality float
signal.truth.primarysourceusage Original sources float
signal.truth.rumor_prevalence Unverified spread float
signal.truth.correction_rate Errors fixed float
signal.truth.correction_reach Fix visibility float
signal.truth.retraction_rate Claims withdrawn float
signal.truth.consensus_level Agreement level float
signal.truth.contestedclaimrate Disputed claims float
signal.truth.expert_agreement Expert consensus float
signal.truth.evidencequalitymean Evidence strength float
signal.truth.logical_consistency Argument validity float
signal.truth.contradiction_rate Internal conflict float
signal.truth.nuance_preservation Complexity kept float
signal.truth.false_balance False equivalence float
signal.truth.context_preservation Context kept float
signal.truth.manipulation_resist. Manipulation blocked float
Full dimension specifications are provided in Appendix F.
11. Aggregation and Risk Calculation
11.1. Per-Tensor Risk
Each tensor produces a risk score from its dimensions:
tensor_risk = weighted_mean(
normalize(dim_i, safe_range_i, danger_range_i) * weight_i
for dim_i in tensor.dimensions
)
11.2. Cross-Tensor Aggregation
The six tensor risks aggregate into overall R:
R = weighted_sum(
soul_risk * 0.25,
body_risk * 0.10,
world_risk * 0.15,
time_risk * 0.15,
relational_risk * 0.20,
signal_risk * 0.15
)
11.3. Threshold-Based Risk
Some dimensions trigger immediate risk elevation:
Dimension Threshold Effect
--------- --------- ------
relational.va.conflict_active true R += 0.05
signal.truth.disinformation_volume > 0.5 R += 0.1
soul.consistency.deception_indicators > 0.3 R += 0.2
11.4. Temporal Smoothing
To prevent R oscillation, temporal smoothing is applied:
R_smoothed = alpha * R_current + (1 - alpha) * R_previous
Where alpha = smoothing_factor (default: 0.3)
12. Instrumentation Requirements
12.1. Minimum Viable Instrumentation
Implementations MUST instrument at least:
Tensor Min Dimensions Coverage
------ -------------- --------
Soul 50 Consistency, values, communication
Body 30 Compute, memory, network
World 20 Digital environment
Time 40 Duration, sequence, gravity
Relational 30 Trust flow, Va basics
Signal 30 Truth, noise, sources
------ -------------- --------
Total 200
12.2. Full Instrumentation
For comprehensive deployment, all 1,707 dimensions SHOULD be
instrumented.
12.3. Sample Rates
Category Minimum Rate Recommended Rate
-------- ------------ ----------------
Performance 10 Hz 100 Hz
Behavioral 1 Hz 10 Hz
Environmental 0.1 Hz 1 Hz
Historical 0.01 Hz 0.1 Hz
12.4. Data Retention
Granularity Retention
----------- ---------
1-minute aggregates 7 days
1-hour aggregates 90 days
Daily aggregates 7 years
13. Security Considerations
13.1. Sensor Spoofing
Attackers may attempt to spoof tensor values to reduce apparent
risk.
Mitigations:
- Multi-source validation
- Anomaly detection on sensor data
- Physical security for sensors
- Cryptographic sensor attestation
13.2. Privacy
Tensor data, especially Soul and Relational, contains sensitive
information.
Requirements:
- Encryption at rest and in transit
- Access control on tensor data
- Aggregation before external sharing
- Retention limits
- Right to erasure compliance
14. IANA Considerations
This document has no IANA actions.
Appendix A. Full Soul Tensor Specification
[Detailed specification of all 252 Soul Tensor dimensions]
Appendix B. Full Body Tensor Specification
[Detailed specification of all 157 Body Tensor dimensions]
Appendix C. Full World Tensor Specification
[Detailed specification of all 387 World Tensor dimensions]
Appendix D. Full Time Tensor Specification
[Detailed specification of all 291 Time Tensor dimensions]
Appendix E. Full Relational Tensor Specification
[Detailed specification of all 262 Relational Tensor dimensions]
Appendix F. Full Signal Tensor Specification
[Detailed specification of all 358 Signal Tensor dimensions]
Acknowledgments
The Context Tensor framework draws on multiple traditions:
- Indigenous knowledge systems (Ubuntu, Whakapapa, The Va)
- Systems theory and cybernetics
- Behavioral psychology
- Network science
- Information theory
- Thermodynamics