大尺度脑仿真实验室 Digital Brain Laboratory
Official@chaobrain
Digital Brain Laboratory provides BrainX, a JAX-based ecosystem for multiscale brain simulation spanning spiking networks, Hodgkin-Huxley cells, and neural-mass models.
Agent Skills by 大尺度脑仿真实验室 Digital Brain Laboratory
Showing 13 vetted skills indexed across 1 GitHub repositories.
chaobrainbrainmass
Simulate, couple, and fit neural-mass models of regional and whole-brain population dynamics.
chaobrainbrainunit
Attach physical units to numerical values and enforce dimensional safety in JAX-based neuroscience computations.
chaobrainbrainevent
Implements event-driven spike communication and plasticity operators for BrainX spiking neural networks.
chaobrainbrainpy-state
Build and train point-neuron spiking networks with BrainPy-State projections and dynamics.
chaobrainbraincell
Build and simulate conductance-based neuron models with Hodgkin-Huxley channels and multicompartment morphology.
chaobrainbrainstate
Build stateful BrainState modules with State-aware JAX transforms and simulation environments.
chaobrainbrainx-general-guard
Routes BrainX modeling tasks to scale-appropriate package skills and enforces BrainX-native implementation.
chaobrainbrainx-install
Installs, upgrades, and uninstalls the BrainX Python meta-package with device-targeted verification.
chaobrainbraintrace
Train recurrent and spiking neural networks with memory-efficient eligibility-trace online learning.
chaobrainbrainx-visualization
Create diagnostic and publication-ready scientific figures from BrainX simulation and analysis results.
chaobrainbrainx-acceleration
Audits and refactors BrainX/BrainState simulation code into state-aware transform patterns for performance.
chaobrainbrainx-modeling-loop
Orchestrates end-to-end BrainX brain-modeling projects from specification through Codex review.
chaobrainbio-neuro-lit
Review biology and neuroscience literature using Europe PMC, Full-Text Resolver, Exa, and DeepXiv.
Frequently Asked Questions About 大尺度脑仿真实验室 Digital Brain Laboratory
FAQPage SchemaWhat tasks can I accomplish with chaobrain's BrainX skills?▼
You can build multiscale brain simulations: point-neuron spiking networks with BrainPy-State, conductance-based Hodgkin-Huxley cells with BrainCell, and neural-mass whole-brain dynamics with BrainMass. Supporting skills cover event-driven synaptic plasticity, unit-safe math, memory-efficient training, GPU acceleration, and scientific figure generation.
Who should use the BrainX skill collection?▼
Computational neuroscientists and research engineers modeling brain dynamics at cellular, circuit, or whole-brain scale. It suits users running spiking-network training, multiscale simulations, parameter sweeps, or neuroscience literature reviews who need dimensionally safe, JAX-accelerated implementations rather than generic machine-learning pipelines.
How do I install and set up the BrainX environment?▼
Use the brainx-install skill, which first checks whether the BrainX meta-package exists in the intended Python environment. It then handles installation, setup, upgrade, downgrade, version pinning, migration, uninstallation, and cleanup, ensuring the correct BrainX stack is present before modeling begins.
How does BrainX handle training memory limits in spiking models?▼
The BrainTrace skill replaces BPTT's sequence-length-dependent computation graph with eligibility traces, enabling memory-efficient temporal training on long sequences. It addresses BPTT out-of-memory failures and activation-memory reduction through estimator selection and compilation, but is not intended for general speed optimization or offline BPTT.
What is the recommended end-to-end modeling process in BrainX?▼
Start with brainx-general-guard to identify modeling scales, then follow brainx-modeling-loop: write a NeuroSpecification.md, research mechanisms via bio-neuro-lit with Europe PMC, implement with the relevant scale skills, accelerate via brainx-acceleration, pass Codex MCP review, and visualize results last.