702 items · Page 9 of 15
Create a free account to keep your reading organized, join thoughtful discussions, and get more from every chapter.
Free to join · Takes less than a minute
Why
I created this space so readers can learn together, ask questions, and make sense of difficult ideas.
QLoRA combines 4-bit quantization with low-rank adapters to fine-tune large language models on limited hardware while preserving model quality.
Implement content safety and moderation in AI agents, including system-level instructions, output filtering, pattern blocking, graceful refusals.
Covers SARIMA (Seasonal AutoRegressive Integrated Moving Average) for forecasting time series with seasonal patterns.
Whisper applies encoder-decoder transformers to multilingual speech recognition, translation, and language identification across noisy audio conditions.
Covers exponential smoothing for time series forecasting, including simple, double (Holt's), and triple (Holt-Winters) methods.
Flamingo connects frozen vision and language models through gated cross-attention, supporting few-shot image and video tasks without task-specific fine-tuning.
Covers LLaMA's architectural choices including RMSNorm, SwiGLU, and RoPE, plus training data strategies.
Covers Google's PaLM, the 540 billion parameter language model that demonstrated advance capabilities in complex reasoning, multilingual understanding.
Examines Qwen's architectural innovations including GQA, SwiGLU activation, and multilingual tokenization.
Explains how the unigram LM tokenizer trains via the EM algorithm, uses Viterbi decoding for segmentation, and enables subword regularization.
Covers Prophet time series forecasting including additive decomposition, trend modeling, seasonal patterns, and holiday effects.
Examines Mistral 7B's architectural innovations including sliding window attention, grouped query attention, and rolling buffer KV cache.
Covers HELM (Holistic Evaluation of Language Models), the early evaluation framework.
Covers GQA, the attention mechanism behind LLaMA 2 and Mistral. Topics include KV head sharing, memory savings, implementation, and quality tradeoffs.
Build agentic workflows with LangChain and LangGraph using state graphs, tool routing, persistence, retries, and multi-step control.
Explains how Multi-Query Attention reduces KV cache memory by sharing keys and values across attention heads, enabling efficient long-context inference.
Explains why traditional word-level approaches fail with diverse text, from OOV words to morphological complexity.
Covers K-means clustering from mathematical foundations to practical implementation. Topics include the algorithm, initialization strategies.
Covers multi-vector retrieval systems introduced in 2021. Explains how token-level contextualized embeddings enabled fine-grained matching.
Create and use test cases to evaluate AI agent performance. Build broad test suites, track results over time.
Examines Microsoft's Phi model family and how textbook-quality training data enables small models to match larger competitors.
Explains how WordPiece differs from BPE by using a likelihood objective for merges. Topics include the PMI score, greedy tokenization, ## prefix notation.
Examines LLaMA's core architectural components: pre-norm with RMSNorm for stable training, SwiGLU feed-forward networks for expressive computation.
Chain-of-thought prompting elicits intermediate reasoning steps from language models. Covers zero-shot and few-shot methods, benefits, and limits.
Define clear, measurable success criteria for AI agents including correctness, reliability, efficiency, safety.
Repetition, frequency, and presence penalties alter token probabilities to reduce loops. Covers n-gram blocking, tuning, and effects on generated text.
Covers t-SNE (t-Distributed Stochastic Neighbor Embedding), including mathematical foundations, probability distributions, KL divergence optimization.
Constrained decoding uses token masks and grammars to produce valid JSON, SQL, and regular-expression matches from language models.
Covers the 2021 Foundation Models Report published by Stanford's CRFM. Explains how this influential report formally defined foundation models.
Covers the mathematical foundations of LLM fine-tuning with clear explanations and minimal prerequisites.
Covers mechanics of autoregressive generation in transformers, including the generation loop, KV caching for efficiency, stopping criteria.
Explains how nucleus sampling dynamically selects tokens based on cumulative probability, solving top-k limitations for coherent and creative text generation.
Covers LIME (Local Interpretable Model-Agnostic Explanations), including mathematical foundations, implementation strategies.
Covers Mixture of Experts (MoE) architectures, including routing mechanisms, load balancing, emergent specialization.
Top-k sampling limits each generation step to the k most probable tokens. Covers probability renormalization, diversity, failure modes, and tuning.
Explains how large language models learn new tasks from prompt demonstrations without weight updates.
Covers OpenAI's InstructGPT research from 2022, including the three-stage RLHF training process, supervised fine-tuning, reward modeling.
Explains how multiple AI agents collaborate through specialization, parallel processing, and coordination.
Explains how temperature scaling reshapes probability distributions during text generation, with mathematical foundations, implementation details.
Covers UMAP dimensionality reduction, including mathematical foundations, fuzzy simplicial sets, manifold learning.
Explains how ELECTRA achieves BERT-level performance with 1/4 the compute by detecting replaced tokens instead of predicting masked ones.
Covers EleutherAI's The Pile, the early 825GB open-source dataset that broadened access to high-quality training data for large language models.
Covers choosing the right approach for your LLM project: using pre-trained models as-is, enhancing them with context injection and RAG.
Examines GPT-2's architecture, model sizes, WebText training, and zero-shot capabilities that changed language modeling through scale.
Covers the core concepts of LLM workflows: connecting language models to tools, handling responses, and building intelligent systems.
Covers BERT fine-tuning for downstream NLP tasks. Topics include task-specific heads, hyperparameter tuning, and strategies to prevent catastrophic forgetting.
Covers Principal Component Analysis, including mathematical foundations, eigenvalue decomposition.
Covers Dense Passage Retrieval (DPR) and Retrieval-Augmented Generation (RAG), the 2020 innovations.
Create a free account to keep your reading organized, join thoughtful discussions, and get more from every chapter.
Free to join · Takes less than a minute
Why
I created this space so readers can learn together, ask questions, and make sense of difficult ideas.