Primers: deep, visual notes you can keep.
Short, carefully made PDF guides on one topic each, in the KinetixAI style: the story, the math, and the code. Buy once, yours to keep.
A whole research project on eight hand-drawn pages. Scan the first two free.
Attention & Transformers
$9From dot-product attention to full transformer blocks, with the math and runnable code.
Types of RAG: GraphRAG, Self-RAG & Corrective RAG
$9The RAG family tree: when plain retrieval breaks, and what GraphRAG, Self-RAG and corrective RAG each fix.
Agentic AI
$12Agents, tools, memory and orchestration, and how MCP and A2A fit together.
Diffusion Models
$12How denoising diffusion works, from the forward process to samplers and guidance.
Core AI: The Foundations Map
$9The one-page map of how AI, ML, deep learning and the rest of this catalog fit together, before you pick a track.
Classical AI: Search, Logic & Planning
$9The pre-neural canon: state-space search, constraint satisfaction, planning and knowledge representation.
Classical ML: Linear Models to Gradient Boosting
$9Regression, trees, boosting and SVMs: the models that still win on tabular data, with the math laid bare.
Neural Network Training: Optimizers & Scaling Laws
$12SGD to Adam, regularization, and the scaling laws that decide how big a model is worth training.
Maths for AI: Linear Algebra, Probability & Calculus
$9Just the math every other Primer assumes you have: vectors and matrices, probability, and the calculus behind backprop.
Optimization & Decision Science
$12Convex optimization, constrained problems, and the decision theory under reinforcement learning and bandits.
Advanced DSA for ML & Systems Interviews
$9The data structures and algorithms that actually come up in ML and infra interviews, not generic leetcode.
Computer Vision: Convolutions to Vision Transformers
$12CNNs, object detection, segmentation, and how vision transformers changed the field.
NLP & NLU: Tokenizers, Embeddings & Language Understanding
$9From tokenization and embeddings to how a model actually understands, not just predicts, language.
Research RAG: Building a System That Actually Grounds Answers
$12Chunking, embeddings, rerankers and evals: the hands-on build, not just the theory behind it.
Agent Factory: Multi-Agent Systems That Ship
$14Moving from one agent to a team of them: orchestration, hand-offs, shared memory and failure containment.
GPU & CUDA: How Your Model Actually Runs on Silicon
$12The GPU memory hierarchy, kernels and parallelism that decide why one model is fast and another is not.
GPU Research: Quantization, Kernel Fusion & Inference
$14The research-side GPU story: quantization, kernel fusion and the tricks behind fast, cheap inference.
MLOps & DLOps: Shipping Models to Production
$12Versioning, CI/CD for models, monitoring and rollback, the SRE playbook applied to ML and deep learning.
Advanced Systems for AI: Distributed Training & Serving
$14Sharding, parallelism strategies and serving infrastructure for models too big for one machine.
AI Security: Prompt Injection, Jailbreaks & Red-Teaming
$12How models actually get attacked, and the defenses and red-teaming practices that hold up.
Paper Notes: Attention Is All You Need
$9The Transformer: sequence modelling with attention alone, no recurrence and no convolution.
Vaswani, Shazeer, Parmar, Uszkoreit, Jones, Gomez, Kaiser, Polosukhin · NeurIPS 2017 · arXiv:1706.03762
Paper Notes: FlashAttention
$12Exact attention, made IO-aware so it fits in fast GPU memory, with large speedups and linear memory.
Dao, Fu, Ermon, Rudra, Ré · NeurIPS 2022 · arXiv:2205.14135
Paper Notes: TriCalRAG
$12A reproducible on-prem benchmark for LLM root-cause analysis in AIOps, from the KinetixAI team at IIT Jodhpur.
Patel, Mohanty, Chaudhary (IIT Jodhpur) · Preprint, 2026 · arXiv:2609.14762
Paper Notes: TriShieldRAG
$12A three-ring defense for RAG, and the adaptive attack that slips past it, from the KinetixAI team at IIT Jodhpur.
Mohanty, Patel, Yuvaraj, Chaudhary, Singhania (IIT Jodhpur) · Preprint, 2026 · arXiv:2607.23838
No Primers in this category yet.
Research Papers Primers are notes on one real, cited paper, grounded in the same data as Paper Lab. Scholar and Professional members get a new one free every month.
Attention & Transformers, Types of RAG, Agentic AI and Diffusion Models, at a saving. Save over 25% versus buying separately.
Every Primer in the catalog, current and future, one purchase. The most anyone pays, no matter how many Primers we add later.
Checkout opens at launch. Join free and we will email you when Primers go live.
Primers are delivered as PDF. Checkout and tax are handled by our merchant of record. Scholar members get 20% off, Professional members 30%. See what else each plan unlocks.