Salesforce AI Associate Certification - 3

What is semantic retrieval in the context of LLMs? 

Searching for relevant information in other data sources

What additional protection does the Trust Layer add to customer data? 

Data masking

What are prompt Défense guardrails in LLMs for? 

Decreasing the likelihood of unintended or harmful output

Supervised learning is typically used with what type of data? 

Structured data

What does supervise learning aim to do? 

Verify output with known, expected results

What is the primary function of classifications? 

Categorize data into groups

What are the key features of Einstein Bots in Salesforce's Service Cloud? 

Automatic issue resolution, case field prediction, and AI-driven customer service

What does Einstein Vision for Field Service automate? 

Image classification for quicker issue resolution on-site

What is a bot in the context of AI? 

An application that automates standard tasks

What is one not a concern related to LLMs and AI-generated content? 

Decreasing carbon footprint of AI models

What is the definition of generative AI? 

Technology that creates new content based on a set of data

What is a key difference between generative AI and predictive AI? 

Generative AI focuses on generating content, while predictive AI predicts future events

What is the goal of part of speech tagging in NLP? 

Identifying the underlying grammatical structure of a sentence

Which technique in NLP focuses on reducing words to their root while considering the part of speech? 

Identifying the underlying grammatical structure of a sentence

What does sentiment analysis involve? 

Determining whether a text expresses a positive, negative or neutral sentiment

What is the primary goal of change management in AI deployment? 

Ensuring the success of new initiatives

What is Einstein Next Best Action a part of within the Salesforce ecosystem? 

Einstein Platform

What tasks can an AI solution for sales teams perform?  

Eliminating time-consuming busywork

What are the Five Trusted AI Principles at Salesforce? 

Responsible, Accountable, Transparent, Empowering, Inclusive

What are the guidelines for Trusted Generative AI related to? 

AI bias mitigation

What does a model card provide important details about? 

AI model characteristics

Which Salesforce AI feature aids in reducing data entry and activity logging for sales reps? 

Einstein Automated Contacts

What does Einstein Opportunity Insights provide to sales reps? 

Intelligent updates about opportunities

What are the key principles outlined in the guidelines for Trusted Generative AI? 

Accuracy, Safety, Transparency, Empowerment, Sustainability

What occurs in overfitting when creating predictive models? 

Memorizing unnecessary details

How does underfitting affect statistical algorithms in modeling? 

Inability to capture underlying patterns

What defines raw data in the context of data analysis? 

Data yet to be processed or modified

What defines Einstein Knowledge within Salesforce? 

A solution providing AI-guided article and solution suggestions

What distinguishes Einstein Vision from Einstein Prediction? 

Einstein Vision handles image recognition, while Einstein Prediction focuses on forecasting

Which Salesforce feature highlights the synergy between AI and human capabilities? 

Empowering

What term describes the challenge faced by businesses in identifying AI-generated profiles that

interact with real users, making it challenging to detect bot networks? 

User Spoofing

What advantage does producing multiple versions of an output provide for generative AI? 

Assisting in choosing the best output version

In what way can generative AI assist when the product description is available? 

Adjusting the tone and voice of the existing description

Which aspects are crucial considerations in AI implementation? 

Data Quality, Data Cleansing and Data Governance

What poses challenges for humans in interpreting AI decisions? 

The complexity of algorithms and reliance on large datasets

What ethical challenges are commonly associated with AI? 

Human bias in algorithms and lack of transparency in decision processes

What crucial role does considering privacy concerns play in handling AI and CRM data? 

Ensures compliance with laws and regulations

To enforce the requirement of having a phone number and pin code for each new contact, which feature should the data quality expert at Salesforcegeek Inc. Use? 

Validation rule

Which practice aligns with Salesforce's Trusted AI Principle of Inclusivity? 

Testing models with diverse datasets

What term describes the process of breaking sentences into separate units, often done by splitting words based on spaces in the English language? 

Tokenization

#salesforce #ai #testing #generativeai #einstein #certification

Salesforce AI Associate Certification - 2

What is it called when you create a prompt that can be used repeatedly for different purposes? 

Prompt template 

How can you prevent LLMs from creating unintended or harmful content? 

Adding specific direction 

What is the process called when data is pulled from your organization to be included in the prompt? 

Dynamic grounding 

What is checked for in the Einstein Trust Layer when the LLM passes a response back to Salesforce? 

Toxicity and Data Demasking 

What is the term for determining the intention behind words in natural language processing? 

Natural language understanding 

What is an algorithm? 

A set of rules for turning input into output 

What is the primary function of Named Entity Recognition (NER)? 

Labelling sequences of words and identifying important entities like names, dates and times 

How does deep learning draw conclusions from data? 

By connecting the dots between data points 

What is the role of training data in machine learning? 

It helps machines make sense of attributes and characteristics. 

What is the GPT language model, one of the biggest LLMs, known as? 

Generative Pre-trained Transformer 

What is the term for predictions from generative AI that diverge from expected responses based on facts? 

Hallucinations 

When can businesses share proprietary data during the generative AI lifecycle? 

Only when using the data to train foundational models 

How does reinforcement learn work? 

It teaches AI models through trial and error, receiving rewards or corrections based on output 

What does NLP stand for in artificial intelligence? 

Natural Language Processing 

What tasks can NLP help computers perform? 

Translating languages 

Which type of data analytics predicts what is going to happen in the future based on past data? 

Predictive 

What is the technique used to discover how strongly two quantitative variables are related? 

Coorelation 

What does the regression line represent in linear regression? 

The best-fitting straight line through data points 

Which Einstein feature helps sales managers identify at-risk opportunities? 

Einstein Opportunity Scoring 

What does Salesforce Einstein Discovery augment within a no-code-required environment? 

Business intelligence 

How many rows with outcome values does Einstein Discovery need to build a model? 

400 

Which of the following Tableau products is best suited for data cleaning, shaping and combining? 

Tableau Prep Builder 

What is the primary key to successful data exploration? 

Formulating Good Questions 

What is the primary function of Einstein Activity Capture in Salesforce? 

Automatically logging activities related to Salesforce records 

What does data-centric AI necessitate for its development? 

Use of only high-quality data 

What is the primary purpose of data preparation in AI implementation? 

Capturing underlying patterns 

How many observations does Einstein Discovery support in datasets? 

Up to 20 million 

What does a data source contain for exploration and understanding? 

Raw data 

Which statement accurately describes a characteristic of continuous variables? 

They have unbroken values 

What does a CRM Analytics dataset for Einstein Discovery require? 

At least 3 columns with one outcome plus two predictor variables 

What is the primary objective of Generative AI? 

Generating content and solutions based on the patterns it learns from existing data or information 

What is the primary objective of integrating AI with Salesforce for Sales and Marketing teams? 

Improving decision-making and efficiency 

What concern arises when AI models for image generation replicate styles from publicly available data without variation or curation? 

Plagiarism 

What process utilizes customer data and machine learning to determine the probability of contacts engaging with emails and marketing campaigns? 

Engagement scoring 

Which primary AI capabilities are predominantly offered by Salesforce? 

Predictive, Generative and Analytic AI 

What is the primary objective of Einstein Lead Scoring within the Sales processes? 

To prioritize sales efforts based on lead conversion 

When integrating AI into their CRM, what is the initial step Salesforcegeek Inc. should take to prepare their data? 

Determine data availability 

If Salesforcegeek Inc. is utilizing Einstein for predictions and encounters inaccurate results, what could be a potential cause? 

Poor data quality 

To address customer complaints regarding excessive sales calls, which data quality aspect should Salesforcegeek Inc. assess to reduce communication inefficiencies? 

Data duplication 

What AI technique instructs a model to optimize outcomes by receiving feedback based on its performance from an algorithm? 

Reinforcement learning 

#salesforce #ai #einstein #generativeai #testing

Salesforce AI Associate Certification - 1

 1. What can large language models (LLMs) help businesses and teams with? 

    Language-based tasks 

2. To get high-quality output from LLMs, what should you add to your prompts? 

    Concrete information and clear constraints 

3. How can you control the variability of LLMs' responses? 

    Adjust the temperature settings 

4. What is "prompt injection" in generative AI? 

    A type of attack on LLMs 

5. What does "zero data retention" mean in the context of Salesforce and external API-based LLMs? 

    Not retaining any customer data outside of Salesforce 

6. What happens to the prompt and responses once they are sent to OpenAI in the Salesforce context? 

    OpenAI forgets the data as soon as the response is sent back to Salesforce 

7. What is natural language processing (NLP) focused on in AI? 

    Understanding how words are used together 

8. What does NLU stand for in the context of communication between people and machines? 

    Natural Language Understanding 

9. How is NLP different from NLU? 

    NLP is about understanding what humans mean when they speak naturally, while NLU handles communication between people and machines. 

10. What is the key advantage of using Einstein Bots in Salesforce? 

    They collect and qualify information using natural language understanding (NLU). 

11. Why is it important for chatbots to integrate with customer data and business processes? 

    To provide a seamless experience and escalate to an agent if necessary 

12. What is generative AI capable of producing? 

    Unique text, images and sounds 

13. What is the primary role of predictive AI? 

    Spotting patterns in data 

14. What is the primary function of a generator in AI? 

    To create new content from a request or input 

15. What does prompt engineering involve? 

    Carefully crafting or choosing the input (prompt) for machine learning models 

16. Why is context analysis important in natural language understanding? 

    To make sense of a piece of text based on surrounding information 

17. What are the three key analysis techniques used in extracting valuable insights from text and speech data? 

    Sentiment analysis, intent analysis, and context analysis 

18. What is the primary goal of Diagnostic data analytics? 

    To understand why something happened 

19. What does Einstein Activity Capture enable sales reps to do in Salesforce? 

    Automatically add activities to related records 

20. What is the primary purpose of Einstein Lead Scoring? 

    Analysing lead conversion history 

21. What does Einstein Forecasting provide predictions about? 

    Sales team opportunities 

22. How many report rows does Einstein Discovery for Reports analyze? 

    500,000 

23. What is the minimum number of rows needed to train a model for predictions and improvements? 

    400 

24. How should the risk of bias in demographic targeting be mitigated according to the content? 

    By using interest and intent-based targeting 

25. What does a model card provide insights into for users? 

    AI model characteristics and limitations 

26. What is the essence of Human-Centered Design (HCD)? 

    Resolving customer problems 

27. What is the most critical action for successful generative AI implementation according to the content? 

    Managing Data 

28. Which term refers to data that hasn't been processed in any way? 

    Raw Data 

29. What is the primary focus of data literacy? 

    Data communication and exploration 

30. How do discrete variables differ from continuous variables? 

    Discrete variables have separate and distinct values 

31. What type of AI promises to create new content and generate solutions based on the existing content? 

    Generative AI 

32. What principles does Salesforce outline for ensuring trustworthy development of Generative AI? 

    Accuracy, Safety, Honesty, Empowerment, Sustainability 

33. What potential risk is associated with real-time personalization in Marketing content within Salesforce? 

    Security compromises like Data Breach 

34. What role can Generative AI play in assisting Salesforgeek Inc. when developing a new marketing campaign? 

    Generating ideas based on market trends and company goals 

35. What's a significant advantage of using AI for code generation? 

    Allowing developers to focus on innovative solutions while reducing time spent on repetitive tasks 

36. Which AI discipline significantly contributes to Salesforce's predictive text and speech recognition functionalities? 

    Natural Language Processing 

37. For enhancing Sales and customer support functions, which specific AI features should Salesforcegeek Inc. Consider? 

    Einstein Lead Scoring and Case Classification 

38. What strategy can users employ to minimize biases while training an AI system? 

    Using diverse training data 

39. What is the probable effect of utilizing high-quality data on customer relationships? 

    Enhanced trust and satisfaction 

40. What term refers to the process of formulating a question precisely to obtain a specific desired answer? 

    Prompt engineering  #salesforce #ai #generativeai #einstein #testing #llms