HPE2-B08 Sample Questions & Answers
Core AI concepts take the single biggest share, alongside positioning HPE's AI lineup by a customer's maturity and workloads, the infrastructure and software behind Private Cloud AI with NVIDIA, plus how big each configuration size runs.
Launch the full HPE2-B08 simulator →Showing 6 of 12 free samples.
- Question 1IntermediateSelect 2
Recognize fundamental AI concepts · Large language models (LLM)
When discussing the underlying technology of modern Large Language Models (LLMs) with a technical audience, an AI architect must accurately describe the components that enable these models to process and generate human-like text. Which TWO of the following are foundational architectural elements or mechanisms primarily responsible for the capabilities of modern LLMs like GPT or Llama? (Select TWO)
Show answer & explanation
Correct answers: A, D
Modern LLMs are built on the Transformer architecture (introduced in the 'Attention Is All You Need' paper), which allows for parallel processing of sequence data. The self-attention mechanism within transformers enables the model to weigh the importance of different words in a sentence relative to each other, granting deep contextual understanding.
Modern LLMs are built on the Transformer architecture (introduced in the 'Attention Is All You Need' paper), which allows for parallel processing of sequence data. The self-attention mechanism within transformers enables the model to weigh the importance of different words in a sentence relative to each other, granting deep contextual understanding.
- Question 2Beginner
Recognize fundamental AI concepts · Generative AI
True or False: In the context of artificial intelligence, Generative AI refers exclusively to models that are designed to classify existing data into predefined categories, such as identifying whether an email is spam or not spam.
Show answer & explanation
Correct answer: B
False. Models that classify data into predefined categories (like spam detection) are known as Discriminative AI. Generative AI, by contrast, is designed to create new, original content (text, images, code, audio) based on learned patterns from its training data.
- Question 3Advanced
Recognize fundamental AI concepts · Retrieval Augmented Generation (RAG)
An AI engineer is evaluating different embedding models to power the vector database in a Retrieval Augmented Generation (RAG) pipeline. The engineer notices that when users ask conceptually similar questions using completely different vocabulary, the system fails to retrieve the correct documents. What is the most likely technical cause of this issue?
Show answer & explanation
Correct answer: C
In a RAG architecture, the embedding model converts text into high-dimensional vectors. If the system fails to match conceptually similar queries that use different vocabulary, the embedding model is likely performing poorly at semantic representation (acting more like a legacy keyword/lexical search). High-quality embeddings place semantically similar concepts near each other in vector space, regardless of the exact words used.
- Question 4Beginner
Recognize fundamental AI concepts · Core AI/ML/DL terminology
When explaining deep learning to a customer new to AI, which structural characteristic distinguishes a Deep Learning (DL) neural network from a simpler Machine Learning (ML) algorithm?
Show answer & explanation
Correct answer: C
The 'deep' in Deep Learning refers to the depth of the artificial neural network structure—specifically, the presence of multiple hidden layers between the input and output layers. These multiple layers allow the network to automatically learn and extract complex, hierarchical features from raw, unstructured data (like images or text) without requiring manual feature engineering.
- Question 5Intermediate
Recognize fundamental AI concepts · Large language models (LLM)
A developer is building an application that interfaces with a large language model. They encounter an error stating 'Maximum token limit exceeded'. Which fundamental limitation of LLM architecture does this error directly relate to?
Show answer & explanation
Correct answer: C
LLMs process text in chunks called tokens. Every LLM has a fixed 'context window' (e.g., 8k, 32k, or 128k tokens) which is the absolute maximum amount of text (input prompt + generated output) the model can hold in its working memory at one time. Exceeding this limit results in errors or the model 'forgetting' earlier parts of the conversation.
- Question 6Intermediate
Recognize fundamental AI concepts · Large language models (LLM)
When distinguishing between the phases of developing a Large Language Model, which phase requires the most massive amounts of unlabelled data and compute power to teach the model general language patterns, grammar, and world facts?
Show answer & explanation
Correct answer: A
Pre-training is the initial and most resource-intensive phase of LLM creation. During this phase, the model consumes massive datasets of unlabelled text (like web scrapes, books, and articles) using self-supervised learning to predict the next word. This establishes its foundational understanding of language. Fine-tuning and alignment happen later using much smaller, curated datasets.
Ready for the real thing?
The full HPE2-B08 simulator has every exam-style question, timed mode, and instant scoring.