Infosys Certified Generative AI Testers -Intermediate
Practice with real exam-pattern questions for Infosys Certified Generative AI Testers -Intermediate. Each question includes a detailed explanation to help you understand the concept, not just memorise the answer. Try 10 questions free — no login required.
Full question bank for this exam + 1,357+ others. Cancel anytime.
Join Premium10 Infosys Certified Generative AI Testers -Intermediate practice questions with answers
Real Lex exam-pattern multiple-choice questions for the Infosys Certified Generative AI Testers -Intermediate certification. Each question includes the correct answer. The full question bank is available to Premium members.
- Question 1
Your team is using an open-source test dataset but finds that many test cases are outdated, unstructured, and inconsistent. What should be done before using them?
- ✓Clean and normalize test cases, remove duplicates, and standardize formatsCorrect
- BUse them directly without modification
- CDelete test cases that don't match your application’s domain
- DMigrate them to a Curated Test Set
- Question 2
In an e-commerce search engine, a user searches for "Apple laptop", but the top results include Apple fruit-related items. What is the best approach to resolve this ambiguity?
- ✓Use Named Entity Recognition (NER) to detect brand vs. common noun usageCorrect
- BApply standard keyword-based search
- CRely on user click-through data to re-rank results
- DIncrease the weight of semantic embeddings in ranking
- Question 3
A law firm’s internal semantic search system fails to retrieve cases when lawyers search using industry-specific jargon. What is the most effective solution?
- ✓Train a domain-specific embedding model (e.g., fine-tune BERT on legal documents)Correct
- BUse a general-purpose embedding model like Word2Vec
- CReduce stopword filtering to keep legal terms intact
- DRely only on keyword-based search with exact matches
- Question 4
A legal search engine expands a query for "Intellectual Property Rights" (IPR) to include synonyms like "Patent Law" and "Copyright Law", but users report that irrelevant documents appear frequently. Which technique can best balance precision and recall?
- ✓Use a hybrid search approach combining semantic and keyword-based retrievalCorrect
- BDisable query expansion and rely only on exact matching
- CIncrease the weight of keyword frequency in ranking
- DManually curate synonym mappings to limit expansion
- Question 5
Your company is developing security test automation scripts and needs real-world security test data from open-source repositories. Which dataset is most relevant?
- ✓OWASP Vulnerable Web Applications DatabaseCorrect
- BMNIST (Handwritten digits dataset)
- CImageNet (Object classification dataset)
- DUCI Machine Learning Repository (General ML datasets)
- Question 6
A resume-matching system consistently ranks candidates from certain universities higher, even when other resumes have better experience matches. What is the most likely cause of this bias?
- ✓Word embeddings learned biases from training dataCorrect
- BThe system relies too much on keyword frequency
- CQuery expansion is overly broad
- DThe model ignores semantic meaning in ranking
- Question 7
When a test automation engineer wants to reuse an existing test case for a new feature, at which stage would they typically interact with the semantic search engine?
- ✓During the test planning phase, before any test cases are written.Correct
- BWhile the automated tests are being executed.
- CDuring the test design and development phase.
- DAfter the test results have been analyzed.
- Question 8
To ensure the semantic search engine remains accurate and up-to-date as the test suite and data evolve, which of the following processes is MOST important to implement?
- ✓Regularly recompiling the code of all test scripts.Correct
- BAutomatically backing up all configuration files daily.
- CEstablishing a process for updating the test data inventory and semantic annotations whenever new test cases are added or existing data requirements change.
- DPeriodically retraining any underlying NLP models on the latest code and documentation.
- Question 9
When implementing semantic search in regression test selection, what is a key challenge?
- ✓Handling synonyms and context variations in defect descriptionsCorrect
- BDatabase size limitations
- CNetwork latency
- DProgramming language incompatibilities
- Question 10
During the indexing stage, what is the primary goal when using semantic embeddings for test cases?
- ✓To count the number of keywords in each test case.Correct
- BTo store the exact text of each test case
- CTo capture the meaning and context of each test case in a vector representation.
- DTo organize test cases alphabetically.
Pay once. Clear every cert this year.
One subscription, full Telegram channel access, every PDF posted during your membership.
- Full access to all 1,357+ certifications
- Monthly updated question banks
- Telegram private channel access
- Cancel anytime
- Everything in Monthly
- Save ₹2,100 vs monthly billing
- Priority answer key requests
- Best for increasing DQ score fast
- Everything in Quarterly
- Lifetime channel access — no renewals
- All future certifications included
- Priority response from admin team
Common questions, straight answers.
A monthly-updated Telegram channel where we post real exam-pattern question banks and detailed answer keys for 1,357+ Infosys Lex certifications. You join once, you get every PDF posted during your membership.
Right after payment on our Graphy page, you'll receive a private invite link to the Telegram channel. Access is instant — usually under 30 seconds.
We compile question banks from the actual Lex test pattern, sourced and verified by 180K+ community members who've recently cleared these exams. Match rate is consistently 85–95%.
Every single month. When Infosys rolls out new versions of certifications, we post updated dumps within 7–10 days. You'll see channel activity weekly.
Clearing certifications is one of the highest-weighted DQ factors. Members typically clear 3–5 certifications in their first 3 months, which moves DQ scores up by a full band.