Topology Search
The simulator moves through a space of possible graphs, comparing structures rather than assuming a fixed ideal form.
Adaptive Graph Systems Simulator
EvolutionLab
EvolutionLab is an experimental simulator for studying how graph structures evolve under utility, mutation, decay, and adaptive selection. It turns network formation into a laboratory for topology, clustering, efficiency, and emergent organisation.
Rather than simulating biology directly, it explores evolutionary behaviour in graph space, revealing which motifs survive, which collapse, and which repeatedly reappear as stable high-utility forms.
Graphs
Networks grow, shrink, reconnect, and reorganise under structural and utility constraints.
Utility
Each topology is evaluated through a weighted objective balancing information, integration, and efficiency.
Motifs
Runs often rediscover compact clustered structures, making recurring graph motifs visible.
History
Milestones, utility traces, and graph states can be compared step by step across experiments.
What EvolutionLab Models
EvolutionLab explores adaptive structure in graph space. It asks how networks behave when links decay, mutations occur, and the system continually selects topologies that better satisfy its utility function.
Emergent Organisation
Clustering, path efficiency, density, and connectivity emerge from repeated updates instead of being prescribed in advance.
Structural Selection
Some motifs recur because they perform well under the objective, acting like attractors in topology space.
The Evolution Model
EvolutionLab treats graphs as evolving entities. Nodes, edges, clustering, and component structure are measured over time, while a utility function rewards efficient and coherent topologies under changing constraints.
Core Utility Form
U = αI + γΦ + δE − βK
- I — information or organisational contribution
- Φ — structural integration / coherence
- E — viability or efficiency term
- K — cost, decay, or complexity penalty
Key Graph Readouts
- node count and edge count
- average path length
- average clustering
- modularity
- largest component size
- utility trajectory over time
Key Parameters
The simulation can be tuned to favour denser graphs, sparser exploration, stronger decay, or more aggressive mutation.
Utility Weights
- alpha
- beta
- gamma
- delta
Evolution Controls
- decay rate
- mutation samples
- pruning thresholds
- run length / steps
Experiment Design
- preset configuration
- seed control
- multiple runs
- milestone and history tracking
Research Use
EvolutionLab is the structural layer of the Blue Whale platform. It complements molecular and agent simulations by showing how adaptive selection operates over network topology rather than physical position or direct symbolic behaviour.
Useful Questions
- Which graph motifs maximise utility under sustained decay?
- When do clustered structures outperform sparse efficient networks?
- How stable are recovered motifs across seeds and parameter sweeps?
- Does the system favour integration over modularity under current weights?
Why It Matters
EvolutionLab makes topology selection measurable. It reveals how network forms emerge, break, recover, and scale, turning graph evolution into something you can inspect as a repeatable experimental process.
Run a topology experiment
Launch EvolutionLab, sweep parameters, and compare how utility shapes clustering, path length, density, and recurring network motifs.