Learn AI &amp; ML the Easy way

Modern ML

Attention

Focus a token and watch how strongly it draws on every other one, in the graph and in the matrix at once. Then watch the weights get spent: the token’s output is the blend they buy, and the temperature decides whether it says anything.

Foundations

Bias-Variance Tradeoff

Sweep a polynomial from a flat underfit to a wild overfit and watch validation error bottom out and turn back up. Then change how much data the model gets, and watch the same degree go from reckless to reasonable.

Evaluation & RL

Threshold, Confusion Matrix & ROC

Drag one threshold and watch the confusion matrix, precision, recall and the ROC point all move together. Then push it to either extreme and see one metric go perfect while the other collapses.

Modern ML

Transformer Decoding

A language model hands you a distribution, not a word. Step through generation and watch greedy decoding walk itself into a loop, then use temperature and truncation to get variety without nonsense.

Supervised Learning

Gradient Descent

Drag the data, drag the line, then watch an optimiser find the fit for you — one real gradient step at a time. Push the learning rate too far and watch it fly apart.

Visuals

Core ML

Classical ML, deep learning, optimisation, and neural network foundations — taught through numbered SVG sequences

Computer Vision

Classical image processing, geometry, detection, and vision architectures — taught through numbered SVG sequences

Concepts

Gen AI

Transformers, LLMs, RAG, Agents, and the generative AI stack

Core ML

Classical ML, deep learning, training patterns, and foundations

Data Concepts

Data engineering, pipelines, and data-centric AI

Math

Linear algebra, calculus, probability for ML

NLP Concepts

Classic NLP pipeline, POS tagging, NER, sentiment analysis, and text processing

Papers

Attention Is All You Need

The landmark 2017 paper that introduced the Transformer architecture, replacing RNNs with self-attention and launching the modern era of large language models.

LLaMA 1

LLaMA: Open and Efficient Foundation Language Models

The 2023 paper from Meta that introduced the first highly capable, open-weights foundation model, sparking the open-source LLM revolution.

Llama 3

The Llama 3 Herd of Models

The 2024 Meta technical report detailing a massive scale-up in data and compute, proving that dense models can reach frontier-level capabilities through sheer scale and data quality.

Llama 2

Llama 2: Open Foundation and Fine-Tuned Chat Models

The 2023 paper from Meta that introduced a commercially viable, chat-tuned foundation model, detailing the immense effort required for RLHF and safety alignment.

GPT-4

GPT-4 Technical Report

The 2023 technical report from OpenAI detailing the model that defined the frontier of AI capabilities, demonstrating human-level performance on professional benchmarks.

Interview Questions

"End-to-End Generative AI System Design"

Quick answer

An end-to-end design splits into an offline data pipeline for embedding knowledge, an online inference pipeline for generating answers via RAG, and an asynchronous feedback loop for evaluation and continuous tuning.

"Optimizing LLM Applications for Latency & Cost"

Quick answer

Optimization relies on aggressive prompt caching, semantic routing, moving simpler tasks to smaller/cheaper models, and using streaming to improve perceived latency.

"Grounding in GenAI: Improving LLM Reliability"

Quick answer

Grounding anchors an LLM's response to an external source of truth, such as a database or search engine, forcing it to synthesize factual information rather than relying solely on its internal, potentially hallucinated weights.

"Vector Embeddings & Semantic Search in GenAI"

Quick answer

Embeddings are numerical vector representations of data that capture semantic meaning. By calculating the distance between these vectors, systems can retrieve documents that mean the same thing, even if they use entirely different words.