65 posts · page 2 of 2
Learn how self-consistency prompting samples multiple reasoning paths and aggregates answers to improve accuracy, with hands-on examples and trade-offs.
Write system prompts that steer model behavior reliably: role, format, constraints, refusals, and evaluation patterns that actually work.
Practical prompt engineering for building software with LLMs: structure, few-shot, chain-of-thought, role messages, and what actually moves quality.
Explore Tree of Thought prompting, which lets LLMs branch, evaluate, and backtrack through reasoning steps to solve problems chain-of-thought cannot.
Compare fixed-size, sentence, semantic, and structural chunking for retrieval augmented generation and pick the right one for your corpus.
An overview of document loaders in RAG pipelines, covering common formats, libraries, and how to choose the right loader for your data.
Graph RAG combines knowledge graphs with retrieval augmented generation to handle multi-hop questions and complex reasoning.
Learn how Hypothetical Document Embeddings (HyDE) improve RAG recall by embedding a generated answer instead of the raw query, with examples and trade-offs.
A practical comparison of RAG and fine-tuning, with guidance on when to choose each, and when to combine them in production systems.
A grounded tour of prompt engineering techniques: clear instructions, structured output, few-shot examples, chain-of-thought, role prompts, and evaluation habits that keep prompts honest.
Walk through the transformer architecture that powers modern LLMs: tokens, embeddings, self-attention, multi-head attention, feed-forward layers, residuals, and the path from input to output.
Why vibes do not scale: building golden datasets, exact-match vs LLM-as-judge scoring, A/B comparing prompts and models, regression suites, and the observability you need to ship safely.
How LLMs call functions: defining tools with JSON schema, the request → tool-call → response loop, common patterns like search and database queries, and the failure modes that bite in production.
How to turn prompts into reusable templates: variables, system messages as contracts, few-shot structure, JSON output schemas, role prompting, and when patterns help versus hurt.
A clear introduction to retrieval-augmented generation — why LLMs don't know your data, how embeddings and vector search solve it, and when RAG beats fine-tuning.
A clear, honest introduction to large language models — tokens, next-token prediction, training vs inference, context windows, why hallucinations happen, and when LLMs are the right tool.
A practical guide to prompting LLMs — system vs user prompts, clarity over cleverness, few-shot examples, structured JSON output, chain-of-thought, and eval-driven iteration.