Skip to content

Lambenthan/empiricalwiki / .claude/skills/prefill

prefill

Verified

用领域基础知识填充 wiki/foundations/,避免后续 /ingest 为教科书材料创建重复的 concept 页面

FileKindSize
SKILL.mdskill md6.0 KB
foundations-catalog.yamlconfig3.4 KB

foundations-catalog.yaml

View on GitHub
domains:
  general:
    - slug: gradient-descent
      title: "Gradient Descent"
      summary: "一阶优化算法家族,通过沿损失函数梯度的反方向迭代更新参数"
    - slug: backpropagation
      title: "Backpropagation"
      summary: "利用链式法则高效计算神经网络中各参数梯度的算法"
    - slug: regularization
      title: "Regularization"
      summary: "防止过拟合的技术家族,包括 L1/L2 正则化、dropout、早停等"
    - slug: cross-validation
      title: "Cross-Validation"
      summary: "将数据划分为训练集和验证集以评估模型泛化能力的方法"
    - slug: optimization
      title: "Optimization"
      summary: "机器学习中最小化损失函数的方法论,包括 SGD、Adam 等优化器"
  NLP:
    - slug: transformer
      title: "Transformer"
      summary: "基于自注意力机制的序列建模架构,取代 RNN 成为 NLP 主流"
    - slug: attention-mechanism
      title: "Attention Mechanism"
      summary: "允许模型动态关注输入不同部分的机制"
    - slug: recurrent-neural-networks
      title: "Recurrent Neural Networks"
      summary: "通过隐藏状态处理序列数据的神经网络架构,已被 Transformer 大幅取代"
    - slug: word-embeddings
      title: "Word Embeddings"
      summary: "将离散词语映射到连续向量空间的表示学习方法"
    - slug: sequence-to-sequence
      title: "Sequence-to-Sequence Models"
      summary: "编码器-解码器架构,用于变长输入到变长输出的映射"
    - slug: language-model
      title: "Language Model"
      summary: "学习文本概率分布的模型,是 GPT/BERT 等预训练模型的基础"
    - slug: tokenization
      title: "Tokenization"
      summary: "将文本切分为模型可处理的基本单元的方法"
    - slug: pre-training
      title: "Pre-training"
      summary: "在大规模无标注数据上学习通用表示的训练范式"
  CV:
    - slug: convolutional-neural-networks
      title: "Convolutional Neural Networks"
      summary: "利用卷积操作提取空间特征的神经网络架构"
    - slug: residual-connections
      title: "Residual Connections"
      summary: "通过跳跃连接缓解深层网络梯度消失的技术"
    - slug: batch-normalization
      title: "Batch Normalization"
      summary: "通过规范化中间层激活值加速训练的技术"
    - slug: generative-adversarial-networks
      title: "Generative Adversarial Networks"
      summary: "生成器和判别器对抗训练的生成模型框架"
    - slug: data-augmentation
      title: "Data Augmentation"
      summary: "通过对训练数据施加变换来增加数据多样性的技术"
    - slug: image-classification
      title: "Image Classification"
      summary: "将图像分配到预定义类别的视觉任务"
  RL:
    - slug: markov-decision-process
      title: "Markov Decision Process"
      summary: "强化学习的数学框架,定义状态、动作、转移和奖励"
    - slug: policy-gradient
      title: "Policy Gradient Methods"
      summary: "直接优化策略函数的强化学习方法族"
    - slug: q-learning
      title: "Q-Learning"
      summary: "学习状态-动作价值函数的无模型强化学习算法"
    - slug: reward-shaping
      title: "Reward Shaping"
      summary: "通过设计辅助奖励信号引导强化学习agent学习的技术"
slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack."Stars: 198kUpdated Claude CodeCodexVerifiedInstall
ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read.Stars: 194kUpdated Claude CodeCodexVerifiedInstall
ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report.Stars: 194kUpdated Claude CodeCodexVerifiedInstall
ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt.Stars: 194kUpdated Claude CodeCodexVerifiedInstall