Sachin Kottarathodi's Space

Category: tech

GenAI Agents

Posted:   |   Categories: tech   |   Tags: Agents, GenAI, LLMs

What is a Generative AI Agent? A GenAI agent is an application that achieves goals by observing the world and acting upon it using tools. It is autonomous, capable of acting independently of human intervention. It reasons, makes decisions, and proactively chooses steps to achieve its goal — even without explicit instructions. Focus is on the specific type of agents powered by Generative AI models. Core Components (Cognitive Architecture) Model – The foundation (e. Read more...

Embedding & Vector Stores: Notes

Posted:   |   Categories: tech   |   Tags: Embeddings, GenAI, LLMs, Vector Stores

Embedding & Vector Stores What are Embeddings? Low-dimensional numerical representation of almost anything (text, audio, images, video). A lossy compression technique: mapping complex data into simpler space while retaining semantic meaning. Enables pattern recognition — embeddings capture semantic similarity and underlying meaning. Useful for efficient large-scale processing. Think of it as: “encoding meaning in vector form.” Use Cases Retrieval & Recommendation systems. Search engines: precompute embeddings for billions of pages, convert search query to embedding, find nearest neighbours in vector space. Read more...

Gemini Code Snippets

Posted:   |   Categories: tech   |   Tags: evaluation, llm, prompt-engineering

Prompt Engineering Using Gemini API Setup from google import genai from google.genai import types client = genai.Client(api_key=GOOGLE_API_KEY) Basic Usage response = client.models.generate_content( model="gemini-2.0-flash", contents="Explain AI to me like I'm a kid." ) print(response.text) Interactive Chat chat = client.chats.create(model='gemini-2.0-flash', history=[]) response = chat.send_message('Hello! My name is Zlork.') print(response.text) Listing Available Models for model in client.models.list(): print(model.name) Model Output Settings # Max output tokens short_config = types.GenerateContentConfig(max_output_tokens=200) # High temperature for creative responses high_temp_config = types. Read more...

Prompt Engineering Notes

Posted:   |   Categories: tech   |   Tags: genai, llm, prompt-engineering

Prompt Engineering Notes Prompt engineering is the process of designing high-quality prompts to guide LLMs to produce accurate outputs. This involves refining length, structure, tone, and clarity. Model-Specific Setup Choose Your LLM and Configure: Model-specific prompts and capabilities Sampling parameters: Output Length More tokens = more compute cost. Reducing length just truncates output. ReAct model warning: Emits irrelevant tokens post-response. Sampling Controls Parameter Effect Temperature Low = deterministic; High = creative/random Top-K Limit prediction to top K likely tokens Top-P Nucleus sampling = choose from top cumulative probability P Num Tokens Max output length Practical Guidelines: Temperature = 0. Read more...