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 neurrate-neural-semantic-narration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 | neurrate-neural-semantic-narration |
| description | NEURRATOR (神经叙述器) - 从单个神经元活动生成自然语言描述的框架，实现单细胞分辨率视觉语义叙述 |
| version | 1.0.0 |
| category | neuroscience |
| tags | ["neural-encoding","single-neuron","natural-language","clip","multimodal","mouse-visual-cortex"] |
| arxiv | 2606.18667 |
| activation_words | ["神经元叙述","neurator","单细胞分辨率","神经编码","视觉语义","CLIP嵌入","Neuropixel"] |
从单个神经元活动生成自由形式自然语言叙述 - 开创性框架将神经脉冲活动解码为对观看场景的自然语言描述，实现单神经元分辨率级别的语义表征理解。
# 核心架构
class NEURRATOR:
def __init__(self):
self.spike_encoder = SpikeToCLIPEncoder() # 脉冲→CLIP嵌入
self.multimodal_llm = FrozenCLIPLLM() # 多模态语言模型
self.sae_validator = SparseAutoencoder() # SAE验证
def narrate(self, spike_trains, neuron_subset):
# 1. 编码脉冲到CLIP空间
clip_embeddings = self.spike_encoder(spike_trains, neuron_subset)
# 2. LLM生成描述
description = self.multimodal_llm.generate(clip_embeddings)
# 3. SAE验证语义一致性
validated = self.sae_validator.validate(description)
return validated
"神经叙述" (Neurration) - 将细胞身份从分类目标转变为视觉系统的功能性探针，提供神经系统中生物学见解的新单位。