{
  "id": 8152754,
  "title": "LLM - Day 3 - Prompt Eng 2",
  "url": "https://urgent.news/2026/09/18/llm-day-3-prompt-eng-2",
  "topic": "ai",
  "section": "AI",
  "published": "2026-09-18T02:12:30.000Z",
  "source": {
    "name": "Dev.to",
    "slug": "dev-to",
    "url": "https://dev.to/technonotes/llm-day-3-prompt-eng-2-3ffc"
  },
  "original_language": "en",
  "account": "The LLM, or Language Model, is essentially a model file that requires training data to function. When prompted with a query, the LLM generates a response based on its pre-existing knowledge. For instance, if one asks the LLM about the current gold rate, it will not be able to provide an answer directly.\n\nThe LLM, being merely an algorithm with pre-trained data, functions as a file. However, it necessitates tools or programs to execute its actions. These tools are essentially scripts. Therefore, one can invoke these tools using an Agent, which is a part of the REACT framework.\n\nThe REACT framework offers two patterns for interaction: RICE and CRISP. Both patterns involve providing a role, input, constraints, and expectations, followed by asking a query. To create an Agent, one needs to install all the necessary items from the techniques folder and run them using the command 'pip install -r file_name'.\n\nThe Agent requires an .env file containing the API key of Groq, which connects to the cloud-based LLM. This connection is crucial because running the LLM locally is not feasible due to the use of gpt-oss-120b. The framework also supports rendering messages, zero-shot and few-shot examples, Tree of Thought options, Chain of Thought, Role-based, Instruction-based, and Contextual Prompting.\n\nThe choice of model is not standardized and depends on the specific use case. Popular models include Hugging Face, Ollama, Blip for images, and Qwen for text. Temperature plays a role in the model's response, affecting its closeness to provided facts or choices.",
  "summary": "REACT Framework ( Reasoning & Action ) LLM is model or a model file. இப்ப ஒரு LLM Model வந்து என்கிட்ட இருக்கு , அந்த Model வந்து ஏதோ ஒரு set of dataவால train ஆகியிருக்கும் நான் இப்ப query கொடுத்தா அது எனக்கு ஒரு response தரும். இப்ப நான் அது கிட்ட போயிட்டு இன்னைக்கு gold rate எவ்ளோ என்று கேட்டால் அது தருமா ? தராதா ? LLM --> Algorithm and pre-trained data is a Model. Also LLM is a file. If LLM…",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}