Last Chance Black Friday Deal: Take 75% Off! Use coupon code below to secure your exam bundles! - Ends in 0d 00h 00m 00s - Coupon code: my75ex

Home > Snowflake > SnowPro Specialty > GES-C02

GES-C02 SnowPro Specialty: Gen AI Exam Question and Answers

Question # 4

A Gen AI Specialist needs to determine whether feedback from a recent survey was generally positive, negative, or neutral. The Specialist wants the assessment reported as either a floating-point number from –1 to +1 (inclusive) or an object with a categories field.

Which statements will produce the desired results? (Choose two)

A.

SNOWFLAKE.CORTEX.CLASSIFY_TEXT (SURVEY_TEXT)

B.

SNOWFLAKE.CORTEX.COMPLETE (SURVEY_TEXT)

C.

SNOWFLAKE.CORTEX.ENTITY_SENTIMENT (SURVEY_TEXT, [‘positive’, ‘negative’, ‘neutral'])

D.

SNOWFLAKE.CORTEX.FINETUNE (SURVEY_TEXT);

E.

SNOWFLAKE.CORTEX.SENTIMENT (SURVEY_TEXT)

Full Access
Question # 5

Which Snowflake data type is returned by the SNOWFLAKE.CORTEX.EMBED_TEXT_768 function?

A.

VECTOR

B.

VARCHAR

C.

VARIANT

D.

OBJECT

Full Access
Question # 6

A Gen AI Specialist is creating metadata for Snowflake Cortex Analyst by creating a semantic model. The relationship between the PRODUCTS and PRODUCT_SALES tables is defined as:

```yaml

relationships:

- name: sales_to_products

join_type: left_outer

relationship_type: many_to_one

left_table: PRODUCT_SALES

relationship_columns:

- left_column: PRODUCT_ID

right_column: ID

right_table: PRODUCTS

```

Which join will Cortex Analyst generate in a query that uses this model?

```sql

A.

SELECT...

FROM PRODUCT_SALES AS S

LEFT OUTER JOIN PRODUCTS AS P

ON S.PRODUCT_ID = P.ID;

B.

SELECT...

FROM PRODUCT AS P

LEFT OUTER JOIN PRODUCT_SALES AS PS

ON PS.PRODUCT_SALES = P.PRODUCTS;

C.

SELECT...

FROM PRODUCT_SALES AS S

FULL OUTER JOIN PRODUCT AS P

ON S.PRODUCT_ID = P.ID;

D.

SELECT...

FROM PRODUCT_SALES AS S

FULL OUTER JOIN PRODUCT AS P

WHERE P.ID = S.PRODUCT_ID;

```

Full Access
Question # 7

Which Snowflake feature simplifies generating vector embeddings and calculating vector similarities, reducing the number of retrieval steps needed in Retrieval Augmented Generation (RAG) pipelines?

A.

Cortex Analyst

B.

Cortex Search

C.

Cortex Fine-tuning

D.

Cortex COMPLETE function

Full Access
Question # 8

A Gen AI Specialist used the ACCOUNTADMIN role to invoke the ` < model_build_name > !PREDICT` method on a Document AI model build and received this error:

`{ "processingErrors": [ "File extension does not match actual mime type. Mime-Type: application/octet-stream" ] }`

What is the MOST LIKELY cause?

A.

The file URL has expired and must be regenerated.

B.

The documents that the query tried to process do not meet the Document AI requirements.

C.

The SNOWFLAKE_SSE encryption type was not specified when creating the internal stage that will be used to store the documents.

D.

The CREATE SNOWFLAKE.ML.DOCUMENT_INTELLIGENCE privilege was not granted to the ACCOUNTADMIN role.

Full Access
Question # 9

Review this query:

```sql

SELECT SNOWFLAKE.CORTEX.CLASSIFY_TEXT(

'I love morning walks',

[

{

'label': 'travel',

'description': 'Hobbies related to going from one place to another',

'examples': ['I like flying to Europe', 'I love beaches']

},

{

'label': 'cooking',

'examples': ['I like learning about new ingredients', 'You must bring your soul to the recipe', 'Baking is my therapy']

},

{

'label': 'fitness',

'description': 'Hobbies related to being active and healthy'

}

],

{ 'task_description': 'Return a classification of the Hobby identified in the text' }

);

```

What will be the result of this query?

A.

{"label": "travel"}

B.

{"label": "fitness"}

C.

"fitness"

D.

{'label': 'fitness', 'description': 'Hobbies related to being active and healthy'}

Full Access
Question # 10

Which view in the ACCOUNT_USAGE schema can query token costs for Snowflake Cortex LLMs aggregated in one-hour increments?

A.

CORTEX_DOCUMENT_PROCESSING_USAGE_HISTORY

B.

CORTEX_FUNCTIONS_QUERY_USAGE_HISTORY

C.

CORTEX_FUNCTIONS_USAGE_HISTORY

D.

CORTEX_SEARCH_SERVING_USAGE_HISTORY

Full Access
Question # 11

A Gen AI Specialist has one document on which they want to perform completions. The document is too large for the context window of SNOWFLAKE.CORTEX.COMPLETE.

In what order should the functions be applied to resolve the issue?

Full Access
Question # 12

Which SQL functions are designed and optimized to carry out and automate specific routine tasks within Snowflake Cortex? (Choose two)

A.

ANOMALY_DETECTION

B.

CLASSIFICATION

C.

FORECAST

D.

PARSE_DOCUMENT

E.

SENTIMENT

Full Access
Question # 13

Which pair of Snowflake Cortex functions should be used together to determine the semantic similarity between documents? (Choose two)

A.

COMPLETE

B.

SUMMARIZE

C.

EXTRACT_ANSWER

D.

EMBED_TEXT_768

E.

VECTOR_L2_DISTANCE

Full Access
Question # 14

Review the following query:

```sql

SELECT DEMO_DB.DEMO_SCHEMA.EXM_DOC_AI_BUILD!PREDICT(

GET_PRESIGNED_URL(@ < stage_name > , ' < relative_file_path > '),

9

);

```

What output will it produce?

A.

The query will produce structured data from a document stored in a Snowflake internal stage using Document AI with a model-version identifier

B.

The query will produce structured data from a document stored in a Snowflake external stage using Document AI with a model-version identifier

C.

The query will produce semi-structured data from a document stored in a Snowflake internal stage using Document AI with a model-version identifier

D.

The query will produce semi-structured data from a document stored in a Snowflake external stage using Document AI with a model-version identifier

Full Access
Question # 15

When working with a multimodal model, what may the SNOWFLAKE.CORTEX.COMPLETE function be used for? (Choose two)

A.

Adding audio to images

B.

Labeling image files

C.

Completing fragmentary images

D.

Classifying landmarks in images

E.

Replacing extracted objects from images

Full Access
Question # 16

A Gen AI Specialist has deployed a base model as an LLM in a production environment for a proof of concept. The model is larger and more expensive than other base models.

How can the Specialist lower production costs without affecting performance for the limited tasks needed to support the proof of concept?

A.

Use Snowflake Cortex Search to limit the number of context documents.

B.

Use Snowflake Cortex Agents to ensure the base model is concise

C.

Use Snowflake Cortex Fine-tuning on a different base model

D.

Provide a semantic model to ensure the LLM is able to interpret the prompts

Full Access
Question # 17

A user opened the Snowflake Cortex LLM Playground in the AI & ML Studio within Snowsight, selected gemma-7b from the Standard models drop-down list, and typed "What is Gen AI?" into the dialog box.

Which SQL statement will achieve the same result?

A.

SELECT SNOWFLAKE.CORTEX.CLASSIFY_TEXT('gemma-7b', 'What is Gen AI?');

B.

SELECT SNOWFLAKE.CORTEX.COMPLETE('gemma-7b', 'What is Gen AI?');

C.

SELECT SNOWFLAKE.CORTEX.EXTRACT_ANSWER('gemma-7b', 'What is Gen AI?');

D.

SELECT SNOWFLAKE.CORTEX.SUMMARIZE('gemma-7b', 'What is Gen AI?');

Full Access
Question # 18

How does Snowflake Cortex Analyst generate suggested questions? (Choose two)

A.

Cortex Analyst only returns questions that are guaranteed to return results.

B.

Onboarding questions flagged in the Verified Query Repository (VQR) are only shown if the user provides no input.

C.

When no Verified Query Repository (VQR) is present, Cortex Analyst uses the underlying LLMs to generate up to three suggested questions.

D.

When a Verified Query Repository (VQR) is defined, Cortex Analyst suggests up to five questions based on semantic similarity to the user's input.

E.

Cortex Analyst prioritizes queries in the Verified Query Repository (VQR) as suggestions based on the queries that reference the most-frequently used logical columns.

Full Access
Question # 19

Which strategy should be used to reduce latency when using the SNOWFLAKE.CORTEX.COMPLETE function?

A.

Use the REST API when invoking inferences

B.

Use the SQL API when using a Snowpark session object

C.

Set the temperature parameter close to 1.

D.

Pass a JSON schema to the response_format parameter

Full Access
Question # 20

Which action is supported when using Snowflake Document AI?

A.

Extracting an entire table in a single query

B.

Processing up to 1000 documents in a single query

C.

Altering a database or a schema where the model build is located

D.

Supporting multiple users working on the same model build at the same time

Full Access
Question # 21

Which Snowflake Cortex AI observability metrics are used to evaluate LLM-generated responses? (Choose two)

A.

Accuracy

B.

Coherence

C.

Correctness

D.

F1 Score

E.

Precision

Full Access
Question # 22

Which proprietary LLM does Snowflake Document AI use?

A.

Arctic-TILT

B.

deepseek-rl

C.

jamba-instruct

D.

mistral-large2

Full Access
Question # 23

Which privileges are needed to run a model fine-tuning job? (Choose two)

A.

READ or WRITE on the stage that contains the model

B.

USAGE on the database used to query the training and validation data

C.

OPERATE on the database used to query the training and validation data

D.

USAGE on the schema where the fine-tuned model will be saved

E.

CREATE MODEL or OWNERSHIP on the schema where the fine-tuned model will be saved

Full Access
Question # 24

Which components are necessary to deploy and run a container with Snowpark Container Services? (Choose two)

A.

A compute pool to run the container

B.

An external stage linked to Amazon S3

C.

A schedule task assigned to a virtual warehouse

D.

A file format registered for JSON or CSV data ingestion

E.

A service specification that defines the container behavior

Full Access
Question # 25

Which governance principles does Snowflake guarantee when its Gen AI features and the underlying models are used? (Choose two)

A.

Snowflake customer data is never made available to other customers or model developers.

B.

Snowflake enforces Role-Based Access Control (RBAC) to ensure appropriate model access.

C.

Snowflake allows customers to create fine-tuned models that can be shared with other customers.

D.

Snowflake only allows third-party models to access Snowflake accounts using the Snowflake Model Registry.

E.

Snowflake mandates the use of Cortex Guard in the Cortex COMPLETE function to filter unsafe responses from a model.

Full Access