菜单
Dual-timescale memory mechanism in spiking neuron-astrocyte networks. Astrocytes provide slow-timescale modulation complementing fast spiking dynamics, enabling energy-efficient learning of environmental patterns and persistent memory traces.
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/hiyenwong/ai_collection --skill dual-timescale-memory-astrocyte命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
Graph-native Python reimplementation of the Information Dynamics of Music (IDyOM) model that represents predictive memories as explicit graph objects for musical expectation modeling and network analysis.
Physics-aware end-to-end deep reinforcement learning methodology for quadcopter control with actuator dynamics modeling.
Reinforced Dreamer methodology for asymmetric reinforcement learning using latent guidance to improve world model representations and behaviors in model-based RL.
基于 SOC 职业分类
正在显示 SKILL.md
| name | dual-timescale-memory-astrocyte |
| description | Dual-timescale memory mechanism in spiking neuron-astrocyte networks. Astrocytes provide slow-timescale modulation complementing fast spiking dynamics, enabling energy-efficient learning of environmental patterns and persistent memory traces. |
| version | 1.0.0 |
| author | Hermes Agent |
| source_paper | Dual-Timescale Memory in a Spiking Neuron-Astrocyte Network for Energy-Efficient Environment Learning |
| paper_url | https://arxiv.org/abs/2604.15391 |
| date | 2025-06-18T00:00:00.000Z |
| tags | ["spiking-neural-networks","astrocyte","dual-timescale","energy-efficient","neuromodulation","memory","biologically-plausible","environment-learning"] |
This skill provides guidance for implementing dual-timescale memory in spiking neural networks using neuron-astrocyte interactions. Astrocytes — glial cells that form the "tripartite synapse" alongside pre- and postsynaptic neurons — provide slow-timescale neuromodulation that complements the fast spiking dynamics of neurons, enabling energy-efficient learning and persistent memory traces of environmental patterns.
For each neuron pair (i, j) with synapse:
Associate astrocyte A_ij with the synapse
A_ij monitors presynaptic spike activity from neuron i
A_ij releases gliotransmitter affecting synapse (i, j)
Neurons: fast spiking dynamics (ms scale)
Astrocytes: slow Ca²⁺ dynamics (s to min scale)
Synapses: modulated by astrocytic state
At each time step:
1. Neurons process input and generate spikes (fast)
2. Astrocytes integrate presynaptic activity
3. Ca²⁺ dynamics update astrocytic state (slow)
4. Gliotransmitter release modulates synaptic efficacy
5. Modulated synapses affect future neural dynamics
Learning emerges from:
- STDP for fast synaptic changes
- Astrocytic modulation for slow, persistent adjustments
- Homeostatic regulation to maintain network stability
Present environmental stimuli to network
Neurons encode stimuli in spike patterns
Astrocytes integrate activity over time windows
Astrocytic traces represent statistical regularities
Network adapts to environmental structure via dual-timescale plasticity
Energy efficiency achieved through sparse spiking + astrocytic gating
| Decision | Recommendation | Rationale |
|---|---|---|
| Neuron model | LIF or Izhikevich | Balance biological plausibility and computational efficiency |
| Astrocyte model | Simplified Ca²⁺ dynamics | Captures essential slow-timescale behavior |
| Coupling mechanism | Synapse-specific astrocytes | Matches tripartite synapse biology |
| Timescale ratio | ~100-1000x slower for astrocytes | Consistent with experimental measurements |
| Learning rule | STDP + astrocytic modulation | Combines fast and slow plasticity |