Google Professional Machine Learning Engineer Practice Test 2
Question 10
Within a jointly maintained Google Cloud data science workspace, the following condition occurs: A daily feature query scans two years of sales even though it uses only yesterday's records because the filter wraps the partition date in an incompatible expression. Collaborators must avoid expanding data access or breaking experiment reproducibility. What should the engineer inspect before selecting a correction?
- Inspect feature event timestamps, prediction cutoffs, feature joins and any post-cancellation administrative flags.
- Examine the partitioning column, query plan, bytes scanned and predicate form against the table's partition definition.
- Inspect join-key frequency distributions, per-task runtime, shuffle spill and the output cardinality before optimization.
- Restart the runtime, run the notebook top to bottom and compare variable initialization with committed source revision.
Explanation
Correct Answer
B:Examine the partitioning column, query plan, bytes scanned and predicate form against the table's partition definition. This inspects evidence relevant to bigquery partition filter cost optimization. Partition pruning avoids irrelevant I/O when predicates match the storage partitioning scheme. After identifying the cause, the matching correction is: Rewrite the filter to permit effective partition pruning while preserving its requested date boundaries.
Incorrect Answers
A:Inspect feature event timestamps, prediction cutoffs, feature joins and any post-cancellation administrative flags. This examines point-in-time feature leakage prevention rather than the specific mechanism reported here. It is not the decisive first check for bigquery partition filter cost optimization. The relevant evidence is: Examine the partitioning column, query plan, bytes scanned and predicate form against the table's partition definition.
C:Inspect join-key frequency distributions, per-task runtime, shuffle spill and the output cardinality before optimization. This examines managed spark join-key skew repair rather than the specific mechanism reported here. It is not the decisive first check for bigquery partition filter cost optimization. The relevant evidence is: Examine the partitioning column, query plan, bytes scanned and predicate form against the table's partition definition.
D:Restart the runtime, run the notebook top to bottom and compare variable initialization with committed source revision. This examines notebook transient state and execution order rather than the specific mechanism reported here. It is not the decisive first check for bigquery partition filter cost optimization. The relevant evidence is: Examine the partitioning column, query plan, bytes scanned and predicate form against the table's partition definition.
Question 11
During a regulated MLOps pipeline deployment, the following condition occurs: A build pushes a model-serving image into production even after its dependency scan reports a critical vulnerability. The workflow must be repeatable and protect the last approved production model. Which result is the strongest postchange evidence of success?
- No-op periods create no unnecessary models and validated drift events produce candidates requiring approval.
- A deliberately vulnerable test image is blocked from production while clean images pass the same gate.
- An audit can resolve model version to exact source data and processing code without guessing by timestamps.
- A deliberately malformed dataset is rejected before training while healthy fixtures proceed successfully.
Explanation
Correct Answer
B:A deliberately vulnerable test image is blocked from production while clean images pass the same gate. This is specific acceptance evidence for cloud build container security gate. Security scanning provides protection only if the CI/CD pipeline enforces its result before release. It confirms the controlled correction: Fail protected build or release stages when prohibited vulnerabilities exist and require documented exceptions for approved deviations.
Incorrect Answers
A:No-op periods create no unnecessary models and validated drift events produce candidates requiring approval. This result could demonstrate success for data drift triggered retraining policy but not for the stated cloud build container security gate issue. The direct test is: A deliberately vulnerable test image is blocked from production while clean images pass the same gate.
C:An audit can resolve model version to exact source data and processing code without guessing by timestamps. This result could demonstrate success for model feature lineage for audit but not for the stated cloud build container security gate issue. The direct test is: A deliberately vulnerable test image is blocked from production while clean images pass the same gate.
D:A deliberately malformed dataset is rejected before training while healthy fixtures proceed successfully. This result could demonstrate success for data validation threshold gate but not for the stated cloud build container security gate issue. The direct test is: A deliberately vulnerable test image is blocked from production while clean images pass the same gate.
Question 12
Within a jointly maintained Google Cloud data science workspace, the following condition occurs: A team imports a pretrained JAX checkpoint in a notebook but assumes its model weights and device setup will load unchanged in a PyTorch-only runtime. Collaborators must avoid expanding data access or breaking experiment reproducibility. Which underlying engineering principle explains why the original approach fails?
- Notebook execution can retain stale objects; a clean deterministic run detects hidden state dependencies.
- MLOps provenance requires a recorded chain from code and data to model and deployment, not merely a friendly name.
- Framework-native checkpoints and runtime APIs are not automatically interchangeable across ML libraries.
- Long-lived secrets copied into notebooks or Git can be reused outside intended workloads and must be rotated.
Explanation
Correct Answer
C:Framework-native checkpoints and runtime APIs are not automatically interchangeable across ML libraries. This explains the mechanism behind model garden jax accelerator compatibility and why the described failure is possible. A discriminating investigation is: Inspect saved artifact format, framework versions, device placement requirements and the model card's supported execution paths. The corresponding repair is: Choose a compatible Model Garden model/runtime or perform a validated checkpoint conversion before training.
Incorrect Answers
A:Notebook execution can retain stale objects; a clean deterministic run detects hidden state dependencies. This principle describes notebook transient state and execution order, which is not the cause of the observed model garden jax accelerator compatibility failure. The controlling explanation is: Framework-native checkpoints and runtime APIs are not automatically interchangeable across ML libraries.
B:MLOps provenance requires a recorded chain from code and data to model and deployment, not merely a friendly name. This principle describes ml metadata lineage for production model, which is not the cause of the observed model garden jax accelerator compatibility failure. The controlling explanation is: Framework-native checkpoints and runtime APIs are not automatically interchangeable across ML libraries.
D:Long-lived secrets copied into notebooks or Git can be reused outside intended workloads and must be rotated. This principle describes workbench credentials embedded in repository, which is not the cause of the observed model garden jax accelerator compatibility failure. The controlling explanation is: Framework-native checkpoints and runtime APIs are not automatically interchangeable across ML libraries.
Question 13
While reproducing a training orchestration incident in isolated staging, the following condition occurs: The same pipeline is manually edited between staging and production, so a critical feature filter differs across environments without a code review. The team requires an idempotent pipeline behavior and traceable evidence. What should the engineer inspect before selecting a correction?
- Inspect task execution states, failed-job logs, alert routing, pipeline-age metrics and owner escalation settings.
- Inspect component definitions, pipeline package versions, parameters and source revisions used in both environments.
- Inspect model-metric artifacts, comparison rules, registry status and deployment dependencies.
- Inspect ML Metadata execution links, table snapshots, transformations and model version lineage.
Explanation
Correct Answer
B:Inspect component definitions, pipeline package versions, parameters and source revisions used in both environments. This inspects evidence relevant to pipeline environment parameter promotion. Promoting immutable pipeline definitions prevents untracked hand edits from changing processing semantics. After identifying the cause, the matching correction is: Publish one versioned pipeline package and promote it with reviewed environment-specific parameters.
Incorrect Answers
A:Inspect task execution states, failed-job logs, alert routing, pipeline-age metrics and owner escalation settings. This examines pipeline notification alert on silent failure rather than the specific mechanism reported here. It is not the decisive first check for pipeline environment parameter promotion. The relevant evidence is: Inspect component definitions, pipeline package versions, parameters and source revisions used in both environments.
C:Inspect model-metric artifacts, comparison rules, registry status and deployment dependencies. This examines model evaluation gate before deployment rather than the specific mechanism reported here. It is not the decisive first check for pipeline environment parameter promotion. The relevant evidence is: Inspect component definitions, pipeline package versions, parameters and source revisions used in both environments.
D:Inspect ML Metadata execution links, table snapshots, transformations and model version lineage. This examines model feature lineage for audit rather than the specific mechanism reported here. It is not the decisive first check for pipeline environment parameter promotion. The relevant evidence is: Inspect component definitions, pipeline package versions, parameters and source revisions used in both environments.
Question 14
During a controlled replay of the data and model handoff to a second engineering team, the following condition occurs: A team uses manual spreadsheet rows for hyperparameter results and loses connections between training metrics and produced checkpoint artifacts. The handoff must preserve privacy, lineage and testable behavior. Which TWO selections form an appropriate before-and-after verification pair? Select TWO.
- Inspect judge rubric prompts, evaluation dataset balance, repeated ratings, human benchmark and uncertainty in pairwise preferences.
- Repeated controlled runs show bounded expected variance and any real improvement exceeds run-to-run noise.
- A newly started Workbench or Colab Enterprise runtime reproduces the complete experiment without undocumented manual cell order.
- Each candidate's reported metrics resolve to its actual artifact and replayable configuration without manual lookups.
- Inspect pipeline execution IDs, experiment tracking records, artifact URIs and the link between metrics and saved models.
Explanation
Correct Answers
D:Each candidate's reported metrics resolve to its actual artifact and replayable configuration without manual lookups. This is specific acceptance evidence for pipeline experiment versus notebook-only tracking. Trial metrics and artifact provenance must be linked to support valid experiment selection and release audits. It confirms the controlled correction: Use Agent Platform Experiments and pipeline metadata to track each trial's parameters, measurements and output artifacts.
E:Inspect pipeline execution IDs, experiment tracking records, artifact URIs and the link between metrics and saved models. This inspects evidence relevant to pipeline experiment versus notebook-only tracking. Trial metrics and artifact provenance must be linked to support valid experiment selection and release audits. After identifying the cause, the matching correction is: Use Agent Platform Experiments and pipeline metadata to track each trial's parameters, measurements and output artifacts.
Incorrect Answers
A:Inspect judge rubric prompts, evaluation dataset balance, repeated ratings, human benchmark and uncertainty in pairwise preferences. This relates to llm-as-a-judge rubric reliability but does not establish the requested before-and-after evidence. The actual diagnostic is: Inspect pipeline execution IDs, experiment tracking records, artifact URIs and the link between metrics and saved models. The specific acceptance result is: Each candidate's reported metrics resolve to its actual artifact and replayable configuration without manual lookups.
B:Repeated controlled runs show bounded expected variance and any real improvement exceeds run-to-run noise. This relates to experiment random seed and data order variance but does not establish the requested before-and-after evidence. The actual diagnostic is: Inspect pipeline execution IDs, experiment tracking records, artifact URIs and the link between metrics and saved models. The specific acceptance result is: Each candidate's reported metrics resolve to its actual artifact and replayable configuration without manual lookups.
C:A newly started Workbench or Colab Enterprise runtime reproduces the complete experiment without undocumented manual cell order. This relates to notebook transient state and execution order but does not establish the requested before-and-after evidence. The actual diagnostic is: Inspect pipeline execution IDs, experiment tracking records, artifact URIs and the link between metrics and saved models. The specific acceptance result is: Each candidate's reported metrics resolve to its actual artifact and replayable configuration without manual lookups.
Question 15
During a production ML inference release on Google Cloud, the following condition occurs: A new image classifier has only offline test results, yet engineers plan to replace every customer prediction immediately after deployment. The service must meet its agreed quality and latency requirements. Which evidence most directly establishes what is wrong with this AI or ML workflow?
- Compare Artifact Registry image digest, build provenance, runtime dependency inventory and release manifests.
- Examine effective IAM bindings, invocation audit logs and the intended partner service account identity.
- Inspect traffic-splitting capabilities, release guardrails, rollback triggers and representative online metrics.
- Inspect experiment assignment IDs, routing hashes, variant weights and per-user exposure histories.
Explanation
Correct Answer
C:Inspect traffic-splitting capabilities, release guardrails, rollback triggers and representative online metrics. This inspects evidence relevant to canary gradual traffic rollout. Progressive traffic exposure reduces blast radius compared with instant replacement of all production serving. After identifying the cause, the matching correction is: Roll out a small canary percentage first, monitor error and quality guardrails and expand only after acceptance.
Incorrect Answers
A:Compare Artifact Registry image digest, build provenance, runtime dependency inventory and release manifests. This examines model artifact image digest reproducibility rather than the specific mechanism reported here. It is not the decisive first check for canary gradual traffic rollout. The relevant evidence is: Inspect traffic-splitting capabilities, release guardrails, rollback triggers and representative online metrics.
B:Examine effective IAM bindings, invocation audit logs and the intended partner service account identity. This examines public endpoint iam oversharing rather than the specific mechanism reported here. It is not the decisive first check for canary gradual traffic rollout. The relevant evidence is: Inspect traffic-splitting capabilities, release guardrails, rollback triggers and representative online metrics.
D:Inspect experiment assignment IDs, routing hashes, variant weights and per-user exposure histories. This examines a/b assignment consistency for user journeys rather than the specific mechanism reported here. It is not the decisive first check for canary gradual traffic rollout. The relevant evidence is: Inspect traffic-splitting capabilities, release guardrails, rollback triggers and representative online metrics.
Question 16
During an independent acceptance rehearsal of a customer-facing AI capability, the following condition occurs: A commerce analyst uses a binary classifier to predict exact revenue per order even though labels are continuous currency amounts. The owners require measurable quality and controlled operating cost. Which statement identifies the primary reason for this failure?
- Unfocused context consumes tokens and can distract generation; targeted retrieval can reduce cost without discarding needed facts.
- In-database prediction avoids unnecessary bulk data transfer when model and features are already in BigQuery.
- Forecasting requires time-ordered histories and explicitly defined future steps, not shuffled categorical labels.
- Classification labels cannot directly represent unrestricted continuous outcomes; SQL regression models match numeric targets.
Explanation
Correct Answer
D:Classification labels cannot directly represent unrestricted continuous outcomes; SQL regression models match numeric targets. This explains the mechanism behind bigquery ml linear regression for continuous value and why the described failure is possible. A discriminating investigation is: Inspect the target column type, business loss metric, and sample prediction schema alongside model_type in the BigQuery model metadata. The corresponding repair is: Use a BigQuery ML LINEAR_REG or suitable regression model with a numeric target and evaluate continuous-error metrics on held-out records.
Incorrect Answers
A:Unfocused context consumes tokens and can distract generation; targeted retrieval can reduce cost without discarding needed facts. This principle describes gemini prompt context token budget, which is not the cause of the observed bigquery ml linear regression for continuous value failure. The controlling explanation is: Classification labels cannot directly represent unrestricted continuous outcomes; SQL regression models match numeric targets.
B:In-database prediction avoids unnecessary bulk data transfer when model and features are already in BigQuery. This principle describes bigquery ml sql batch prediction integration, which is not the cause of the observed bigquery ml linear regression for continuous value failure. The controlling explanation is: Classification labels cannot directly represent unrestricted continuous outcomes; SQL regression models match numeric targets.
C:Forecasting requires time-ordered histories and explicitly defined future steps, not shuffled categorical labels. This principle describes bigquery arima_plus forecast horizon, which is not the cause of the observed bigquery ml linear regression for continuous value failure. The controlling explanation is: Classification labels cannot directly represent unrestricted continuous outcomes; SQL regression models match numeric targets.
Question 17
At a Google Cloud analytics team preparing a governed pilot, the following condition occurs: An analyst normalizes numerical inputs in a notebook before training but later calls ML.PREDICT on raw table columns, shifting prediction values. The team must satisfy a reproducible low-code AI assessment. Which result is the strongest postchange evidence of success?
- The scheduled report processes the full authorized partition and stores traceable predictions without a local export.
- A controlled raw record yields the same transformed feature representation for evaluation and production SQL inference.
- An unseen labeled cohort receives probabilities for the intended positive class with a calibrated operating threshold.
- The chosen version meets task quality, inference cost and license rules on a documented representative evaluation set.
Explanation
Correct Answer
B:A controlled raw record yields the same transformed feature representation for evaluation and production SQL inference. This is specific acceptance evidence for bigquery ml transform serving parity. Transformations embedded with a BigQuery ML model reduce divergence between fitting and serving features. It confirms the controlled correction: Move deterministic feature transformations into the supported BigQuery ML TRANSFORM workflow and use the same model at inference.
Incorrect Answers
A:The scheduled report processes the full authorized partition and stores traceable predictions without a local export. This result could demonstrate success for bigquery ml sql batch prediction integration but not for the stated bigquery ml transform serving parity issue. The direct test is: A controlled raw record yields the same transformed feature representation for evaluation and production SQL inference.
C:An unseen labeled cohort receives probabilities for the intended positive class with a calibrated operating threshold. This result could demonstrate success for bigquery ml logistic model for conversion probability but not for the stated bigquery ml transform serving parity issue. The direct test is: A controlled raw record yields the same transformed feature representation for evaluation and production SQL inference.
D:The chosen version meets task quality, inference cost and license rules on a documented representative evaluation set. This result could demonstrate success for model garden task and license selection but not for the stated bigquery ml transform serving parity issue. The direct test is: A controlled raw record yields the same transformed feature representation for evaluation and production SQL inference.
Question 18
Within a jointly maintained Google Cloud data science workspace, the following condition occurs: A model evaluation cell appears to pass only because an old variable remains in kernel memory after cells were executed out of order. Collaborators must avoid expanding data access or breaking experiment reproducibility. What should the engineer inspect before selecting a correction?
- Inspect repository history, mounted credentials, authorized identities and any logs of secret access from the shared project.
- Restart the runtime, run the notebook top to bottom and compare variable initialization with committed source revision.
- Inspect judge rubric prompts, evaluation dataset balance, repeated ratings, human benchmark and uncertainty in pairwise preferences.
- Inspect Python runtime versions, pip dependency trees, package hashes and notebook execution metadata for both runs.
Explanation
Correct Answer
B:Restart the runtime, run the notebook top to bottom and compare variable initialization with committed source revision. This inspects evidence relevant to notebook transient state and execution order. Notebook execution can retain stale objects; a clean deterministic run detects hidden state dependencies. After identifying the cause, the matching correction is: Refactor hidden state into explicit inputs and functions and add a clean-kernel end-to-end execution gate.
Incorrect Answers
A:Inspect repository history, mounted credentials, authorized identities and any logs of secret access from the shared project. This examines workbench credentials embedded in repository rather than the specific mechanism reported here. It is not the decisive first check for notebook transient state and execution order. The relevant evidence is: Restart the runtime, run the notebook top to bottom and compare variable initialization with committed source revision.
C:Inspect judge rubric prompts, evaluation dataset balance, repeated ratings, human benchmark and uncertainty in pairwise preferences. This examines llm-as-a-judge rubric reliability rather than the specific mechanism reported here. It is not the decisive first check for notebook transient state and execution order. The relevant evidence is: Restart the runtime, run the notebook top to bottom and compare variable initialization with committed source revision.
D:Inspect Python runtime versions, pip dependency trees, package hashes and notebook execution metadata for both runs. This examines notebook dependency lock reproducibility rather than the specific mechanism reported here. It is not the decisive first check for notebook transient state and execution order. The relevant evidence is: Restart the runtime, run the notebook top to bottom and compare variable initialization with committed source revision.
Question 19
As the team tests a new governed AI release against adversarial and normal data, the following condition occurs: An online classifier uses a 24-hour account activity total while training created a seven-day total with the same column name. The fix must retain legitimate model usefulness and appropriate controls. Which outcome most directly demonstrates a successful, measurable fix?
- The assistant meets groundedness and task quality thresholds with reduced tokens and p95 latency.
- For controlled entities the online and offline features match at equivalent cutoffs and quality recovers.
- A failing canary halts and restores the healthy model before broader traffic exposure.
- Attribution changes are explained or remediated and decision drivers remain stable on known reference cases.
Explanation
Correct Answer
B:For controlled entities the online and offline features match at equivalent cutoffs and quality recovers. This is specific acceptance evidence for training-serving feature skew detection. Identical feature names do not guarantee identical semantics; mismatched windows create training-serving skew. It confirms the controlled correction: Version a single feature contract and add automated parity tests to the preprocessing pipeline and serving workflow.
Incorrect Answers
A:The assistant meets groundedness and task quality thresholds with reduced tokens and p95 latency. This result could demonstrate success for llm latency and cost token regression but not for the stated training-serving feature skew detection issue. The direct test is: For controlled entities the online and offline features match at equivalent cutoffs and quality recovers.
C:A failing canary halts and restores the healthy model before broader traffic exposure. This result could demonstrate success for production incident rollback based on metrics but not for the stated training-serving feature skew detection issue. The direct test is: For controlled entities the online and offline features match at equivalent cutoffs and quality recovers.
D:Attribution changes are explained or remediated and decision drivers remain stable on known reference cases. This result could demonstrate success for feature attribution drift after deployment but not for the stated training-serving feature skew detection issue. The direct test is: For controlled entities the online and offline features match at equivalent cutoffs and quality recovers.
Question 20
During a staged restoration test at a secondary serving deployment, the following condition occurs: A team deploys a model using a mutable storage path named production-latest and cannot reconstruct which evaluation approved its weights. The team must prove an actual user-visible improvement without weakening access controls. Which underlying engineering principle explains why the original approach fails?
- Containerized CPU inference can be appropriate on serverless compute; specialized ML endpoints are not required for every model.
- Batch workloads need throughput and scheduling guarantees, whereas real-time endpoints serve interactive low-latency calls.
- Endpoint invoker permissions should follow least privilege, rather than expanding access for convenience.
- Mutable nicknames do not establish immutable model provenance for controlled production releases.
Explanation
Correct Answer
D:Mutable nicknames do not establish immutable model provenance for controlled production releases. This explains the mechanism behind model registry version provenance and why the described failure is possible. A discriminating investigation is: Review immutable artifact IDs, Model Registry versions, evaluation reports, approvals and deployed revision metadata. The corresponding repair is: Promote an explicitly approved model version from Agent Platform Model Registry and record deployment lineage.
Incorrect Answers
A:Containerized CPU inference can be appropriate on serverless compute; specialized ML endpoints are not required for every model. This principle describes cloud run versus managed inference endpoint, which is not the cause of the observed model registry version provenance failure. The controlling explanation is: Mutable nicknames do not establish immutable model provenance for controlled production releases.
B:Batch workloads need throughput and scheduling guarantees, whereas real-time endpoints serve interactive low-latency calls. This principle describes batch versus online inference schedule, which is not the cause of the observed model registry version provenance failure. The controlling explanation is: Mutable nicknames do not establish immutable model provenance for controlled production releases.
C:Endpoint invoker permissions should follow least privilege, rather than expanding access for convenience. This principle describes public endpoint iam oversharing, which is not the cause of the observed model registry version provenance failure. The controlling explanation is: Mutable nicknames do not establish immutable model provenance for controlled production releases.