1Z0-1127-25 Practice Test 22 — 15 Questions

Question 1

What is the primary purpose of the self-attention mechanism in a Transformer model?

Accepted Answer

To compute a weighted sum of all token representations based on pairwise relevance. Self-attention allows each token to attend to every other token in the sequence, capturing contextual relationships regardless of distance.

Answer

To generate token embeddings in parallel

Answer

To reduce the dimensionality of token embeddings

Answer

To encode positional information of tokens

Question 2

Which prompting technique involves providing the model with a few examples of input-output pairs within the prompt to guide its behavior?

Accepted Answer

Few-shot prompting. Few-shot prompting is the technique of including a small number of input-output examples directly in the prompt so the model can infer the desired pattern, format, or reasoning style. By showing the model concrete demonstrations, it generalizes the task without any fine-tuning. This is distinct from zero-shot (no examples) and from chain-of-thought (which elicits step-by-step reasoning rather than demonstrating input-output pairs).

Answer

Zero-shot prompting

Answer

Self-consistency prompting

Answer

Chain-of-thought prompting

Question 3

Which of the following sampling strategies selects tokens based on a cumulative probability threshold from the highest probability tokens?

Accepted Answer

Top-p (nucleus) sampling. Top-p (nucleus) sampling cuts off the tail of the probability distribution where cumulative probability exceeds p, allowing dynamic vocabulary size.

Answer

Top-k sampling

Answer

Greedy decoding

Answer

Temperature sampling

Question 4

A data scientist needs to fine-tune a model using OCI Generative AI. They have prepared a dataset in JSONL format with prompt/completion pairs. The fine-tuning job is configured with the T-Few technique. What is a key characteristic of T-Few fine-tuning?

Accepted Answer

It modifies a small subset of parameters using lightweight adapter layers. T-Few is a parameter-efficient fine-tuning method that updates only a small fraction of model parameters, making it faster and more resource-efficient than full fine-tuning. It does not train all parameters, add new layers, or require unlabeled data.

Answer

It requires unlabeled data for unsupervised pre-training before fine-tuning

Answer

It updates all model parameters, requiring significant compute resources

Answer

It only trains a new classification head on top of a frozen base model

Question 5

A developer is using the OCI Generative AI Chat API with Cohere Command R+ to build a multi-turn conversational agent. They want the agent to always respond in a formal tone, regardless of the user's phrasing. Which parameter should they set in the API request to achieve this consistently?

Accepted Answer

Use the 'system' parameter (or preamble_override) to provide a system message like 'You are a formal assistant'. A system message or preamble override sets the overall behavior and tone of the assistant for the entire conversation. Temperature controls randomness; max tokens limits length; stop sequences end generation — none are suitable for defining a persistent tone.

Answer

Set the 'stop_sequences' parameter to include periods

Answer

Set the 'temperature' parameter to 0.1

Answer

Set the 'max_tokens' parameter to 200

Question 6

An organization needs to fine-tune a Cohere Command R model for a custom domain. They have prepared a dataset in JSONL format. Which component of the fine-tuning job specifies the base model and the training dataset location?

Accepted Answer

The fine-tuning job creation request body. The fine-tuning job creation request includes parameters such as 'baseModelId' and 'trainingDataset' (pointing to the dataset in Object Storage). The model deployment endpoint and cluster are separate steps.

Answer

The dedicated AI cluster settings

Answer

The T-Few configuration file

Answer

The model deployment endpoint configuration

Question 7

A data scientist needs to compare two summaries generated by different models against a reference summary. Which metric focuses on recall of n-grams and is commonly used for summarization evaluation?

Accepted Answer

ROUGE. ROUGE (Recall-Oriented Understudy for Gisting Evaluation) is the correct metric because it focuses on recall of n-grams, measuring how many of the reference summary's n-grams appear in the generated summary. This makes it the standard metric for summarization evaluation, as it directly assesses content coverage.

Answer

BLEU

Answer

Perplexity

Answer

BERTScore

Question 8

An AI developer is building a document Q&A application using LangChain and OCI Generative AI. They need to split large PDF documents into smaller chunks before embedding. Which text splitter should they use to ensure splits respect sentence boundaries while also controlling chunk size?

Accepted Answer

RecursiveCharacterTextSplitter. RecursiveCharacterTextSplitter is LangChain's recommended splitter for general text because it recursively splits on a hierarchy of separators (paragraphs, then sentences, then words) to keep semantically related text together while respecting a target chunk size. This makes it ideal for PDF documents where you want sentence-boundary-aware chunks with controlled size.

Answer

WebBaseLoader

Answer

PDFLoader

Answer

TokenTextSplitter

Question 9

An LLM application generates product descriptions. The output is sometimes repetitive (e.g., 'innovative' appears multiple times). Which parameter adjustment is MOST likely to reduce this repetition without harming creativity?

Accepted Answer

Set frequency penalty to a positive value, e.g., 0.3. Frequency penalty penalizes tokens proportionally to how many times they have already appeared in the generated text, directly discouraging the model from repeating the same word (like 'innovative'). A modest positive value such as 0.3 reduces repetition while preserving the model's overall creativity because it does not flatten the probability distribution the way temperature changes do.

Answer

Increase temperature from 0.7 to 1.0

Answer

Decrease max tokens from 200 to 100

Answer

Increase top-k from 40 to 80

Question 10

A developer is using the OCI Generative AI Python SDK. They receive a 400 error 'InvalidParameter'. What is the most likely reason?

Accepted Answer

Invalid model name. The 400 'InvalidParameter' error in the OCI Generative AI Python SDK specifically indicates that the request contained a parameter that is not recognized or is malformed. The most common cause is providing an incorrect or unsupported model name, as the SDK validates model names against a predefined list of available models in the OCI Generative AI service. Other parameter errors, such as invalid temperature or top_p values, would also produce this error, but model name mismatches are the most frequent trigger.

Answer

Exceeded token limit

Answer

Network timeout

Answer

Missing API key

Question 11

You need to convert a set of customer support tickets into vector embeddings for a similarity search application. Which OCI Generative AI model should you use?

Accepted Answer

Cohere Embed (e.g., embed-english-v3.0). Cohere Embed models (such as embed-english-v3.0) are specifically designed to generate vector embeddings from text for semantic search, clustering, and classification. Converting support tickets into embeddings for similarity search is exactly the use case for an embedding model. OCI Generative AI offers Cohere Embed as a managed embedding service.

Answer

Cohere Rerank

Answer

Meta Llama 3

Answer

Cohere Command R

Question 12

A company is using OCI Generative AI Agents to build a customer support assistant. They have uploaded product manuals to OCI Object Storage. Which two components are required to create the agent? (Select TWO)

Accepted Answer

Create a Generative AI Agent using the knowledge base. Option C is correct because OCI Generative AI Agents requires a knowledge base that ingests and indexes the source documents; here the product manuals reside in an Object Storage bucket, so creating a knowledge base from that bucket is a mandatory step. Option A is correct because, once the knowledge base exists, you must create the Generative AI Agent and associate it with that knowledge base so the agent can perform retrieval-augmented generation over the manuals. Option B is not required: InferenceClient is a lower-level SDK construct for calling models directly, not a component needed to build an agent. Option D is wrong because data must be supplied through a knowledge base rather than uploaded directly to the agent. Option E is unnecessary because Dedicated AI Clusters are for dedicated hosting capacity, whereas agents can be created on on-demand infrastructure.

Answer

Create an endpoint via InferenceClient

Answer

Upload data directly to the agent (not via knowledge base)

Answer

Provision a Dedicated AI Cluster

Question 13

A machine learning engineer is evaluating the performance of a translation model using BLEU score. Which THREE statements about BLEU are correct? (Choose three.)

Accepted Answer

BLEU includes a brevity penalty to penalize outputs that are too short. BLEU is a precision-based metric (not recall). It uses modified n-gram precision with a brevity penalty. It correlates reasonably well with human judgment at the corpus level but has known limitations such as not capturing semantic equivalence.

Answer

BLEU measures recall of n-grams by comparing the output to the reference

Answer

BLEU is a recall-oriented metric

Question 14

A developer is debugging a RAG pipeline where the LLM frequently ignores retrieved documents and produces hallucinations. Which THREE factors could contribute to this problem?

Accepted Answer

The prompt does not explicitly instruct the model to base its answer on the provided context. Low chunk overlap can cause loss of context, improper prompt instructions may lead the model to ignore retrieved content, and a small top-K retrieval may miss relevant documents. A large embedding dimension improves retrieval precision, and a high temperature increases randomness but does not cause ignoring documents.

Answer

The generation temperature is set to 0

Answer

The embedding model's output dimension is too large

Question 15

A developer notices that the ConversationalRetrievalChain in their LangChain application is not retaining context from previous turns in the conversation. Which component is most likely missing or misconfigured?

Accepted Answer

A memory component like ConversationBufferMemory. ConversationalRetrievalChain requires a memory object to persist chat history across turns; without it, each invocation is stateless and the chain cannot reformulate follow-up questions using prior context. ConversationBufferMemory (or another BaseMemory implementation) is passed via the memory parameter and is what enables the chain to inject prior turns into the question-condensation step. The other components listed (splitter, retriever, embeddings) affect retrieval quality but do not provide conversational state.

Answer

A document splitter to chunk the history

Answer

A retriever with appropriate search parameters

Answer

An embedding model to vectorize the history