Introduction to Large Language Models 2025/2026 Exam Questions Marking Scheme New Update | A+ Rated

Introduction to Large Language Models
2025/2026 Exam Questions Marking
Scheme New Update | A+ Rated



large language models - 🧠ANSWER ✔✔large, general-purpose language

models can be pre-trained (basic common language problems: text

classification, question answering, document summarization, text

generation) and then fine-tuned (specific problems in different fields)

specific purposes


3 major features of large language models - 🧠ANSWER ✔✔1. large:


-large training dataset

- a large number of parameters: memory and knowledge machine-learned

from training, define the skills of the model




2. general purpose: model sufficient ability to solve a common problem

-commonality of human languages

-resource restriction

COPYRIGHT©JOSHCLAY 2025/2026. YEAR PUBLISHED 2025. COMPANY REGISTRATION NUMBER: 619652435. TERMS OF USE. PRIVACY
STATEMENT. ALL RIGHTS RESERVED
1

, 3. pre-trained and fine-tuned


benefits of using large language models - 🧠ANSWER ✔✔1. a single code

can be used for different tasks




2. the fine-tuning process requires minimal field data

-obtain decent performance even with minimal training

-can be used for few-shot or even zero-shot scenarios




3. the performance of LLM is continuously growing when you add more

data and parameters


few-shot - 🧠ANSWER ✔✔in machine learning, it refers to training a model

with minimal data


zero-shot - 🧠ANSWER ✔✔in Machine learning, implies that a model can

recognize things that have not explicitly been taught in the training before


Pathways Language Model (PaLM) - 🧠ANSWER ✔✔-has 540 billion

parameters

COPYRIGHT©JOSHCLAY 2025/2026. YEAR PUBLISHED 2025. COMPANY REGISTRATION NUMBER: 619652435. TERMS OF USE. PRIVACY
STATEMENT. ALL RIGHTS RESERVED
2