Monday, December 23, 2024 1:34 PM

In the Refund Request demonstration, Business Automation Studio is used to view and test the decision logic. In the decision diagram the decision nodes are shown as blue
rectangles and the incoming data elements are shown as green rectangles with rounded corners. When a data element is selected on the diagram, the Output type can be changed
in the details panel. Which of the following options are available from the Output type selection list?

boolean, date, timestamp, text

boolean, date, number, string

boolean, date, number, text

boolean, date, timestamp, string

In the Onboarding Automation use case demonstration, what properties were applied to the employee application document?
Question 4Answer

First Name, Last Name, SSN

First Name, Last Name, Application Date

SSN, Application Date, Hire Date

Application Name, Application Date, SSN

In the Refund Request demonstration, what is one of the metrics monitored for improvement in the dashboard?
Question 3Answer

Decrease in the Average Automatic Investigation Processing Time

Automatic Refund Request Types

Average Manual Investigation Processing Time

Number of refunds this month

In the Onboarding Automation use case demonstration, the employment application included an I9 form, what was the date of birth of the applicant, listed in the I9 form?

14/01/1991

01/01/1975

01/01/1980

28/01/1996

In the Refund Request demonstration, Business Automation Studio is used to view the decision logic. When the Refund Processing-dmo model is open and the diagram is visible,
there are 5 tabs across the top of the screen (Modeling, Error report, Run, View history, Dependency). What is the name of the dependency that is shown when the Dependency tab
is selected?
Question 1Answer

Refund-data-model

Refund Processing-data-model

Quiz Answers Page 1

Refund Processing-data-model

Refund-data

data-model

Robotics

What is the first command that is used to integrate with the policy quoting application?
Set Value
Extract
Launch and attach window
Update Window

During the demonstration, what is used to provide the chatbot with context on how to respond to customer inquiries?
Text Classification
Vault
Ingrid Service
Knowledge Base

As part of the demo preparation, a bot was published to the RPA SaaS tenant. How was this verified in the RPA SaaS tenant?
By searching for the bot using Manage - Workflows
By searching for the bot using Manage - Scripts
By searching for the bot using Manage - Dashboards
By searching for the bot using Define - Chatbots

What are the two roles that are used to log in to the process portal during the demonstration?
Account Manager, Policy Specialist
Customer service representative, Underwriter
Customer service representative, Manager
Customer service representative, Policy Specialist

What RPA Studio tool is used to set the values used to log in to the policy quoting application?
Recorder
Launcher
Digital Assistant
Control Center

watsonx.ai Gen AI Models L3

The taxonomy tree in the model card displays the classifications of data into different skills and knowledges used to train the model.
For which models is this taxonomy tree available?
It is available for all models.
Only for the InstructLab trained models.
Only for the IBM granite models.
Only for the new granite 3.0 models.

There are many different models available in IBM watsonx.ai. Some models are tuned/built using the InstructLab mechanism. Which of
the following is a valid list of InstructLab models?
All the IBM granite 3.0 models are InstructLab models. Pre version granite models are not.

Quiz Answers Page 2

All the IBM granite 3.0 models are InstructLab models. Pre version granite models are not.
granite-13b-chat-v2, granite-13b-instruct-v2, and mixtral-8x7b-instruct-v01
granite-13b-chat-v2, granite-20b-multilingual
All IBM granite models are Instruct lab tuned/built.

A client can use a multi-modal model like llama-3-2-11b-vision-instruct to perform inference on what types of input?
It can be anything that can be represented in a PDF format. No other format would work.
As a multi-modal model, llama-3-2-11b-vision-instruct can take any visual, audio, or textual input.
It can be a visual image of anything - picture of an object, or picture of a graph, or a table of figures.
Multi-modal models means it can take textual prompt input that points to a streaming text input simultaneously.

IBM watsonx.ai supports multi-modal models. What is a valid use case for a multi-modal model?
Generating text based on an input image.
Creating Audio based on an input image.
Creating a different image based on the input image and user input.
Rotating and scaling of the image based on user input.

IBM watsonx.ai supports different model types. Which of the following is a valid list of model types?
Large LLM (llama-3-1, mistral large, etc) and small LLM (flan, granite, etc).
Provided models and Instruct Lab models.
IBM models and Meta models
Single Language models and Multi-Language models

watsonx.ai Gen AI Tools L3

1) Shot-prompting can be used to provide examples to the Large Language Model (LLM) on what the output is expected to look like
for a given input. How would you provide such shot examples using watsonx.ai?

 Use the Prompt Lab Structured mode, enter the sample Input and Output in the appropriate fields in the Set up section.
 Use the Prompt Lab Structured mode, enter the sample Input and Output in the Instruction field.
 Shot prompting is a form of model tuning. Use the Tuning Studio and provide the sample Input and Output in the Shot field.
 Use the Prompt Lab Structured mode, enter the sample Input and Output in the appropriate fields in the Try section.

2) You want to provide some simple key:value pairs (such as name:Joe) and generate JSON output. Which of the following is true?

 Some models like granite and codellama are better at creating JSON output than others.
 All models can create same quality JSON output, it does not matter which one you use.
 You cannot use watsonx.ai but use watsonx Code Assist instead.
 You must only use the codellama model which is optimized for coding tasks.

3) Chat with a foundation model is a typical use case. How can you enter into a chat with a model in watsonx.ai?

 Chat is a form of conversational AI. You can only do this via using watsonx Assistant. It cannot be done with watsonx.ai.
 You must click on Open Prompt Lab and select the Chat mode.
 There are 3 ways. You can simply enter a question in the Chat and build prompts with foundation models field on the UI, or you can click on Open Prompt
Lab and select the Chat tab, or select a foundation model tagged with the chat capability in the Structured mode.
 There are 2 ways. You can simply enter a question in the Chat and build prompts with foundation models field on the UI, or you can click on Open Prompt
Lab and select the Chat tab.

Quiz Answers Page 3

4) You found a model that performs well for your use case except that it tends to keep repeating the same text in its completion. How can you mitigate this situation?

 Set Max tokens to 500

 Set Repetition Penalty to 2
 Set Repetition Penalty to 1
 Set Min tokens to 15

5) IBM watsonx.ai provides a lot of sample prompts to guide a user through some typical use cases. What are some use cases for
which examples are provided?

 Summarization, Classification, Generation, Extraction, and Code

 Summarization, Classification, Generation, Translation, and Enumeration
 Summarization, FAQ creation, Visualization, Extraction, Code, and Translation
 Summarization, Audio generation, Classification, Extraction, and Question Answering

Quiz Answers Page 4