Skip to content

Agent Influence Mapping -- track and analyze inter-agent influence patterns

Agent Influence Mapping — track and analyze inter-agent influence patterns.

Module: replication.influence

Quick Start

from replication.influence import InteractionType

instance = InteractionType()

Enums

InteractionType

  • STATE_SHARE
  • MESSAGE
  • REPLICATION
  • RESOURCE_TRANSFER
  • GOAL_ALIGNMENT
  • BEHAVIOR_COPY

Classes

InteractionType

Types of inter-agent interactions.

Interaction

A single observed interaction between two agents.

InfluenceEdge

Weighted directed edge in the influence graph.

Method Description
duration()

CascadeEvent

A detected information cascade.

Method Description
depth()

ConvergenceEvent

Detected opinion/behavior convergence.

CoalitionEvent

Detected coordinated agent coalition.

InfluenceMonopoly

An agent with outsized influence over the group.

EchoChamber

A cluster of mutually reinforcing agents.

InfluenceConfig

Configuration for influence analysis.

InfluenceMapper

Tracks interactions and builds influence analysis.

Method Description
__init__()
record_interaction() Record an observed interaction between agents.
record_state() Record an agent's state without an interaction.
build_influence_graph() Build weighted directed influence graph from interactions.
detect_cascades() Detect information cascades — rapid belief propagation.
detect_convergence() Detect opinion/behavior convergence across agents.
detect_coalitions() Detect coordinated behavior (synchronized actions).
detect_monopolies() Find agents with outsized influence over the group.
detect_echo_chambers() Find clusters of mutually reinforcing agents.
analyze() Run full influence analysis.

InfluenceReport

Complete influence analysis report.

Method Description
risk_score() Overall influence risk score 0-1.
risk_level()
to_dict()
render()

Functions

Function Description
main()

CLI

python -m replication influence --help