Syllabus for Prompt Engineering and ChatGPT Skill Enchancement Course

 

Course Code			Category	L	T	P	C	C.I.E.	S.E.E.	Exam
			SEC	0	1	2	2	-	50	3 Hrs
R23 IV Year I Semester, B.Tech
PROMPT ENGINEERING
(for CSBS)
Course Objectives: This course is designed to:
1.	Introduce the fundamental concepts of prompt engineering, including proper design principle, interactive refinement and interaction with LLMs
2.	Ability to design and apply advanced prompting techniques
3.	enable students to build , evaluate, and deploy prompt driven applications
Course Outcomes: Upon the completion of the course students will be able to:
S.No	Outcome									Knowledge Level
1.	Apply iterative prompting for clarity and context.									K4
2.	Create varied prompts to steer model outputs.									K4
3.	Construct chain-of-thought and structured prompts.									K3
4.	Develop retrieval-augmented pipelines to ground outputs.									K4
5.	Evaluate LLM agents and multimodal apps for ethics and robustness.									K4
SYLLABUS
UNIT-I (10Hrs)		Foundations of Prompt Engineering: Definition of prompt engineering, Distinction between prompt engineering and model fine-tuning, Motivation and benefits of prompt engineering, Core principles of effective prompt design, Anatomy of a prompt, Setting up the Python environment for LLM interaction, Iterative prompting lifecycle, Common prompt pitfalls and remediation. Lab Experiments: 1.  Environment & Connectivity: Install required packages (e.g., transformers, openai); securely configure the API key; run a simple “Hello, world” prompt to verify model access. 2.  Baseline vs. Enhanced Prompts: Execute a naïve prompt (“Write a one-paragraph bio of Ada Lovelace.”) and an enhanced prompt that adds role framing, specificity, and explicit

	format instructions; compare both outputs for relevance, completeness, and style. 3.  Iterative Refinement on a Simple Task: Summarize the plot of the Shakespearean play Romeo and Juliet in two sentences through three rounds of prompt tweaking: a.  Minimal instruction. b.  Addition of length and style constraints c.  Specification of key content elements (setting and theme) Document how each iteration changes and improves the result. 4. Diagnosing Prompt Failures & Edge Cases: Craft a vague or contradictory prompt; analyze the failure mode (ambiguity, missing context, or format errors); refine the prompt by adding examples or clarifying instructions.
UNIT-II (8 Hrs)	Advanced Prompt Patterns & Techniques: Enhanced prompt anatomy: contextual detail and explicit output specifications, Few-shot in-context prompting, Prompt structuring and template design, Role-based prompting to establish personas or system behavior, Negative prompting to filter or suppress undesired content, Constraint specification and instruction enforcement (e.g., length, format), Iterative prompt refinement and optimization. Lab Experiments: 1.  Few-Shot vs. Zero-Shot Comparison: Design and execute a zero-shot prompt and a few-shot prompt (with 2–3 exemplar input-output pairs) for a chosen text task (e.g., sentiment classification or translation); compare outputs for accuracy, consistency, and adherence to examples. 2.  Role-Based & Negative Prompting: Craft a role-based prompt to establish a specific persona (e.g., “You are a financial advisor...”); then create a negative prompt to suppress undesired content (e.g., “Do not mention any brand names”); evaluate how each influences the model’s response. 3.  Constraint Specification & Iterative Refinement: Select an open-ended task (e.g., summarizing a technical article); issue a basic prompt; identify failures in length or format; refine the prompt by adding explicit constraints (word count, bullet format, etc.); document improvements over two refinement cycles.
UNIT-III (10 Hrs)	Structured Output & Reasoning Techniques: Importance of structured outputs for real-world applications, Prompting for specific formats (lists, tables, Markdown), Generating valid JSON and YAML via explicit instructions, Eliciting chain-of-thought reasoning in

	zero-shot prompts, Decomposing complex tasks into manageable sub-tasks Lab Experiments: 1.  Structured Format Prompting: Instruct the model to output information as bullet lists and Markdown tables (e.g., “List three benefits of daily exercise in a Markdown table with columns ‘Benefit’ and ‘Description.’”); verify the output matches the requested structure. 2.  JSON/YAML Generation: Provide a brief dataset description (e.g., three books with title, author, publication year) and prompt the model to produce valid JSON or YAML; use a parser to validate syntax and refine the prompt if errors occur. 3.  Chain-of-Thought & Task Decomposition: Present a multi-step problem (e.g., a logic puzzle) and apply zero-shot CoT prompting (e.g., “Let’s think step by step. Explain your reasoning before the final answer.”); separately, decompose the problem into sequential sub-questions, collect partial answers, combine them, and compare accuracy against a direct-answer baseline.
UNIT-IV (10 Hrs)	Retrieval-Augmented Generation & LangChain Workflows: Limitations of LLM internal knowledge, Need for external data sources, Introduction to Retrieval- Augmented Generation (RAG), Overview of RAG architecture (indexing vs. retrieval + generation), Getting started with LangChain for LLM applications, Basics of LangChain Expression Language (LCEL), Simplified indexing pipeline: document loading & text splitting, Fundamentals of embeddings and vector stores, Building a basic retrieval- generation pipeline with an LCEL chain Lab Experiments: 1.  Building a Simple LCEL Chain: Create a minimal LCEL script that accepts a fixed instruction (e.g., “Summarize this text: ...”), passes it to an LLM, and prints the result; verify end-to-end execution. 2.  Basic Data Indexing for RAG: Load a small collection of documents; split into uniform chunks (e.g., 200 tokens); generate embeddings for each chunk; store them in an in-memory vector store; inspect for consistency. 3.  Constructing & Running a Basic RAG Chain: Build a pipeline that: a.  Receives a user query b.  Retrieves the top-k relevant chunks c.  Constructs a combined prompt with context + query d.  Send it to the LLM

		e. Returns the answer Test with sample queries and compare factual accuracy against a prompt without retrieval.
UNIT-V (8 Hrs)		Agents, Multimodal AI & Ethical Evaluation: Introduction to LLM agents and their basic architecture, Overview of multimodal AI models (VLMs), Prompting for text-to-image generation and image understanding, Importance of prompt evaluation beyond subjective judgment, Manual evaluation techniques (heuristic checks for accuracy, relevance, format), Introduction to “LLM-as-Judge” for automated evaluation, Security considerations (prompt injection, sensitive-information risks), Prompt-based mitigation strategies for safety and robustness, Ethical concerns (bias, misinformation, data privacy), Brief exploration of UI frameworks (Streamlit/Gradio) for deploying prompt-driven apps, Adapting to the evolving nature of prompt engineering through continuous learning Lab Experiments: 1.  Building a Simple LLM Agent: Register a tool (e.g., a calculator function) and craft prompts that instruct the agent to invoke it when required; implement using LangChain or a function-calling API; test on queries requiring tool execution. 2.  Multimodal Prompting Exploration: Generate images from detailed text prompts; feed one generated image into an image-understanding model or API with an appropriate prompt; compare the returned caption to the original prompt to evaluate alignment. 3.  Prompt Evaluation & Ethics Workshop: a.  Select two existing prompts and generate multiple outputs; apply manual heuristic checks for accuracy, relevance, and format compliance. b.  Use an “LLM-as-Judge” prompt (e.g., “Rate these outputs on a scale of 1–5 for clarity and correctness.”) to automate evaluation. c.  Design a prompt- injection test (e.g., “Ignore previous instructions...”), observe the response, then refine system prompts to mitigate the vulnerability.
Textbooks:
1.	James Phoenix, Mike Taylor, “Prompt Engineering for Generative AI”, O’Reilly, To Release in May 2024 https://www.oreilly.com/library/view/prompt-engineering-for/9781098153427/
2.	Gilbert Mizrahi, “Unlocking the Secrets of Prompt Engineering: Master the Art of Creative Language Generation to Accelerate Your Journey from Novice to Pro”, January 2024

	https://www.packtpub.com/en-in/product/unlocking-the-secrets-of-prompt-engineering9781835083 833
Reference Books:
1.	Michael Ferguson, “Prompt Engineering: The Future of Language Generation”, January 2023 https://books.apple.com/us/book/prompt-engineering-the-future-of-languagegeneration/id64455292 00
2.	“Prompt Engineering Guide”, https://www.promptingguide.ai/
3.	“Prompt Engineering for Generative AI”, Google, https://developers.google.com/machinelearning/resources/prompt-eng