Skip to main content 홈 크리에이터 hiyenwong ai_collection connectome-wiring-statistical-dynamics-separation
connectome-wiring-statistical-dynamics-separation Separating wiring-specific from statistical control of dynamics in a complete connectome. Analysis of larval Drosophila brain showing coarse statistics set dynamical regime while specific wiring determines activity routing.
설치로 이동 Skills Marketplace 커뮤니티가 만든 AI 스킬을 발견하고 탐색하세요.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
직접 명령은 검토 Prompt를 거치지 않습니다. 실행하기 전에 소스를 확인하세요.
npx skills add https://github.com/hiyenwong/ai_collection --skill connectome-wiring-statistical-dynamics-separation명령은 한 줄로 유지됩니다. 복사하기 전에 가로로 스크롤해 전체 내용을 확인하세요.
로컬 사본을 원하시나요? SkillsMP에서 현재 제공할 수 있는 파일을 다운로드하세요.
Zip 다운로드 다운로드 중... name connectome-wiring-statistical-dynamics-separation description Separating wiring-specific from statistical control of dynamics in a complete connectome. Analysis of larval Drosophila brain showing coarse statistics set dynamical regime while specific wiring determines activity routing. version 1.0.0 category neuroscience arxiv_id 2606.17745 author Stavros Therianos institution Independent Researcher published 2026-06-16T00:00:00.000Z activation_words ["connectome dynamics","wiring statistics","statistical versus specific wiring","larval drosophila","connectome control","network operator gain","mushroom body dynamics","olfactory pathway routing","degree-weight matching","null model hierarchy"] related_skills ["connectome-wiring-statistics-control-dynamics","brain-network-controllability","connectome-constrained-neural-network","connectome-genetic-environmental-architecture","effective-plasticity"]
Separating Wiring-Specific from Statistical Control of Dynamics in a Complete Connectome
Core Innovation
首次在完整连接组 （果蝇幼虫脑）上系统性分离：
粗粒统计特性 ：决定全局动力学 regime（增益、维度、线性度）精细接线模式 ：决定活动传播路径和主导电路
核心发现 : 统计特性设定 regime，精细接线设定几何。
Methodology
Frozen Operator Assay
将连接组作为固定动力学算子 运行：
无单神经元参数调优 : 所有属性归因于接线Rate-based model : 无动作电位、时间常数、突触动力学Spectral radius ρ = 0.99 : Leaky-tanh update
W = connectome_weight_matrix
x_{t+1 } = tanh(ρ * W @ x_t + input )
Null Model Ladder
从最不保留接线到最保留接线：
Level Model Preserves 1 Unstructured Gaussian Nothing 2 Degree + Weight Matched In/out-degree, weight distribution 3 Block-Preserving Rewire + Cell-class architecture 4 Connectome Exact placement
关键 : Level 2-3 保留统计特性但 scramble 接线位置。
Structural-Dynamical Properties
Operator Gain : How much it amplifies inputDimensionality : How many directions surviveMode Leverage : Which neurons shape modesSparse-Input Routing : Where input lands
Key Results
Global Regime: Statistical Control
Degree+Weight matched ensemble reproduces :
Operator gain
Dimensionality
Near-linearity
Implication : 这些全局性质不依赖具体接线，只依赖统计特性。
Pathways: Wiring-Specific Control
Sparse input routing :
Connectome: Activity confined to olfactory pathway
Rewired networks: Activity floods widely
Leading modes concentrated in MB (learning center)
Rewiring distributes modes uniformly
Convergence neurons depleted from driving modes
Cell-Class Architecture Contribution Partial reproduction of confinement :
Block-preserving rewire captures ~50% of routing specificity
Remaining 50% driven by fine synaptic placement
MB-specific wiring concentration :
Not just cell-class labels
Fine-grained synapse placement matters
Null Hierarchy Working: Retraction Example Lateral horn localization claim :
Initial observation: LH concentrates driven-side modes
Size-matched control: Random sets show similar
Result : Claim retracted (false positive)
Anatomical Context
Larval Drosophila Brain
3,013 neurons total111,243 directed connections Strongly connected core : 2,825 neurons, 109,438 synapses536 self-loops
Identified Circuits Circuit Neurons Function Mushroom Body (MB) 231 Learning center Lateral Horn (LH) 201 Olfactory processing Central Complex (CX) 77 Navigation Remaining 2,316 Mixed populations
Input/Output Ports
Afferent : 80 input portsEfferent : 97 output ports
Theoretical Framework
Operator Gain
Dimensionality dim = count(eigenvalues with |λ| > threshold)
Mode Leverage (Adjoint-Side)
mode_leverage = W^T @ dominant_eigenvectors
Sparse-Input Routing
activation_pattern = W^n @ sparse_input
Biological Interpretation
Statistics → Regime Coarse features determine :
Overall responsiveness (gain)
Information capacity (dimensionality)
Linear/nonlinear behavior
Why : All neurons have similar connectivity statistics → similar dynamical role
Wiring → Geometry Specific placement determines :
Which pathways activate
Which circuits dominate dynamics
Where information flows
Why : Exact synapse locations create privileged channels
Mushroom Body Significance
Concentrates leading adjoint modes
Shapes which neurons drive dynamics
Learning center has outsized control
Not cell-class alone : Fine MB-specific wiring critical
Null Model Details
Degree + Weight Matching
for neuron in connectome:
in_degree[n] = count(inputs)
out_degree[n] = count(outputs)
total_weight[n] = sum (connections)
rewire_connections(preserve=in /out_degree, weights)
Block-Preserving Rewire
blocks = [MB, LH, CX, Others]
for block in blocks:
preserve_internal_connections(block)
scramble_between_blocks()
Size-Matched Random Control
circuit = MB neurons
random_set = random_neurons(size=231 )
compare(circuit, random_set)
Experimental Validation
White-Noise Drive
Input through afferent ports
Measure core state evolution
No behavioral/physiological state
Readout Metrics
Operator gain : Signal amplificationDimensionality : Active subspaceMode leverage : Driving neuronsRouting confinement : Pathway specificity
Limitations
What Model Omits
Action potentials
Single-neuron time constants
Synaptic/receptor kinetics
Neuromodulation
Behavioral state
Gap junctions
Interpretation Caution Not physiological simulation :
Structural-dynamical instrument
No odor-evoked activity claims
Wiring-only analysis
Applications
Connectome Analysis
Wiring vs statistics claims : Null model hierarchyCircuit dominance : Mode leverage analysisPathway routing : Sparse-input confinementArchitecture effects : Block-preserving tests
Comparative Connectomics
Cross-species statistics comparison
Developmental wiring changes
Evolutionary wiring optimization
Neural Network Theory
Weight initialization insights
Architectural inductive bias
Connectivity regime analysis
Technical Requirements
Dependencies
NumPy/SciPy (linear algebra)
NetworkX (graph operations)
Matplotlib (visualization)
Hardware
CPU: Network analysis
RAM: ~10GB for full connectome
Key Takeaways
Separation : Statistics set regime, wiring sets geometryNull hierarchy : Different claims need different controlsMB dominance : Learning center shapes dynamicsRetraction example : Null models catch false positivesFrozen operator : Wiring-only attribution
Citation @article{therianos2026connectome,
title={Separating wiring-specific from statistical control of dynamics in a complete connectome},
author={Therianos, Stavros},
journal={arXiv preprint arXiv:2606.17745},
year={2026}
}
Activation : Use when analyzing connectome dynamics, wiring vs statistics separation, network regime analysis, or mushroom body functional dominance in complete connectomes.
