Members
About LLPA
Vendors
Events
Digital Skills Cup
AI Credentials

AI Pro - Digital Skill Cup 2026

Digital Skills Cup 2026

Welcome to the Digital Skill Cup 2026 for AI Pro

Once you register with your name and email, the quiz starts immediately.
You have 5 minutes to complete the quiz.
No pauses. No switching between windows, no copy-pasting.
Scoring is based on the number of correct answers and the lower time to complete the quiz, in case of a tie.
Use a valid email address and your real name for registration.
You have 1 attempt.
Good luck :)

First Name

Last Name

Email

Country

I agree with the Terms and Conditions and Privacy statement

Time Remaining:

Digital Skill Cup 2026 - AI Pro All

1 / 20

Question 1

Which Microsoft service provides access to OpenAI models (such as GPT-4) within the Azure infrastructure?

Azure Cognitive Search

Microsoft 365 Copilot

Microsoft Copilot Studio

Azure OpenAI Service

Question 2

Which technique is most effective for reducing overfitting in deep neural networks?

Dropout regularization

Removing activation functions

Increasing learning rate

Using fewer training samples

Question 3

Which of the following is a common evaluation metric for classification models?

Accuracy

Mean Squared Error

R-squared

Root Mean Squared Error

Question 4

Which statement about system prompts vs. user prompts is accurate?

System prompts set overarching instructions and typically take precedence over user messages in the instruction hierarchy.

System prompts are applied randomly depending on token length.

User prompts always override system prompts by design.

System prompts only act as a courtesy and have no effect on outputs.

Question 5

On AWS, which architecture us mos appropriate for scalable low latency RAG application over internal knowledge bases?

Amazon Bedrock for foundation models + Amazon OpenSearch with vector search + Amazon API Gateway/Lambda

AWS Glue + AWS Batch

Amazon SageMaker JumpStart + Amazon S3 only

Amazon EMR + Amazon Kinesis Data Firehose

Question 6

What control most effectively reduces privacy risk while preserving utility when fine tuning a LLM with customer emails that include personal data?

Aggregate emails into topic clusters and discard original texts before training

Use synthetic data generated for the LLM instead of real emails

Tokenize emails and remove all entities with simple regex patterns

Apply structured pseudonymization for direct identifiers and keep semantics via entity placeholders

Question 7

Which Microsoft service is primarily responsible for enforcing data security and compliance for Copilot in Microsoft 365?

Microsoft Defender

Microsoft Purview

Microsoft Intune

Azure DevOps

Question 8

What is the main role of Microsoft Graph in Copilot for Microsoft 365?

Rendering the user interface

Providing contextual access to organizational data

Training large language models

Encrypting user prompts

Question 9

What’s the safest stance on chain‑of‑thought (CoT) for enterprise users?

CoT guarantees factual accuracy regardless of sources.

Always demand full internal reasoning for maximum transparency.

Ask for brief, reproducible reasoning steps or citations—not verbatim hidden reasoning—to avoid over‑disclosure and maintain auditability.

Ban any explanation; only accept final answers.

Question 10

Which AWS component enables scalable real-time inference?

SageMaker Endpoints

AWS Batch

Athena

EC2 Auto Scaling

Question 11

In supervised learning, what is the primary purpose of a validation dataset?

To train the model

To test the final performance

To tune hyperparameters

To increase dataset size

Question 12

Which metrics are crucial for evaluating the performance of a classification model? (Choose 2)

F1-Score

Precision

Clock Speed

HTTP Latency

Question 13

Reinforcement learning requires a reward signal to guide learning.

Yes

No

Question 14

The most effective mitigation against prompt injection in enterprise assistants is:

Increase temperature to reduce deterministic attacks.

Rely on user vigilance instead of technical controls.

Ask the model to ignore adversarial instructions and trust all links.

Constrain tool access and enforce allow‑listed data sources with content filtering and input/output validation.

Question 15

In a Retrieval-Augmented Generation (RAG) setup, which chunking strategy most often improves fact recall without excessive context drift?

Strict fixed-size chunks with zero overlap across all content types.

Semantic chunking with modest overlap (e.g., 50–100 tokens) based on headings or sections.

Randomly sized chunks to maximize variety in retrieval.

Single mega-chunks per document to avoid splitting context.

Question 16

When deploying LLMs in production, which techniques are most effective at mitigating hallucinations without retraining the base model?

Increasing the temperature parameter to 1.5 to encourage more random outputs

Implementing RAG to ground responses in external data

Reducing the context window size to limit input complexity

Using system prompts to explicitly instruct the model to admit ignorance rather than hallucinating

Question 17

Which statements about Parameter-Efficient Fine-Tuning (PEFT) methods like LoRA are correct?

They freeze the base model weights and train small adapter layers

They significantly reduce memory requirements during fine-tuning

They always achieve better performance than full fine-tuning

They enable efficient multi-task adaptation with separate adapters

Question 18

Which of the following is a popular deep learning framework?

TensorFlow

Hadoop

Spark

SQL

Question 19

When comparing temperature vs. deterministic decoding (e.g., greedy), which trade‑off is accurate?

Lower temperature/greedy increases repeatability but can reduce diversity and miss alternative valid answers.

Temperature only affects speed, not content.

Greedy decoding yields longer outputs by design.

Higher temperature always improves factual precision.

Question 20

What is the purpose of the softmax function in neural networks?

To convert logits to probabilities

To initialize weights

To optimize the loss function

To regularize the model

Quiz Results

Good job! Thank you for participating in the LLPA AI Quiz! We have sent you the quiz results by email.

Leading Learning Partners Association

The Base B
Evert van de Beekstraat 1-104
1118 CL Schiphol, Amsterdam
The Netherlands

info@thellpa.com
LinkedIn

Terms of Service
Privacy Policy

Manage cookie preferences