Tokenization Changes What a Language Model Sees
← Back to Blog & Articles A language model does not receive words directly. It receives token identifiers produced by […]
LLMs Derived, Uncategorized
LLMs Derived, Uncategorized
Why Scaled Dot-Product Attention Divides by Square Root of d
← Back to Blog & Articles The scaling factor in attention is not decorative. It keeps dot products in a […]
The AI Algorithms Compendium, Uncategorized
RAG Is a Pipeline, Not a Single Algorithm
← Back to Blog & Articles Retrieval-augmented generation connects a language model to an external collection. Its quality depends on […]
The AI Algorithms Compendium, Uncategorized
Q-Learning and Policy Gradients Solve Different Problems
← Back to Blog & Articles Q-learning estimates the value of actions; policy-gradient methods adjust a policy directly. Both optimize […]
The AI Algorithms Compendium, Uncategorized
Why Attention Replaced Recurrence in Many Sequence Models
← Back to Blog & Articles Recurrent networks process sequence positions through a chain of hidden states. Attention creates direct, […]
The AI Algorithms Compendium, Uncategorized
Bias and Variance: Diagnosing Model Error
← Back to Blog & Articles Bias and variance describe two different ways a learned model can fail. They are […]
The AI Algorithms Compendium, Uncategorized
Bayesian Updating: From Prior Belief to Posterior Evidence
← Back to Blog & Articles Bayesian inference is a disciplined way to revise uncertainty. It combines what was believed […]
The AI Algorithms Compendium, Uncategorized
BFS, Uniform-Cost Search, and A*: What Changes?
← Back to Blog & Articles BFS, uniform-cost search, and A* can look nearly identical in code. Their difference is […]
Algorithmic Edge, Uncategorized
How to Explain Algorithm Correctness in an Interview
← Back to Blog & Articles A correct program is stronger when you can explain why it must work. A […]
Algorithmic Edge, Uncategorized
Greedy or Dynamic Programming? Test the Irreversible Choice
← Back to Blog & Articles Greedy algorithms commit to a locally attractive choice. Dynamic programming keeps alternatives alive through […]