VISCA converts webpages into a semantic component abstraction and uses it as context for LLM-generated end-to-end tests, reporting 92% average feature coverage on E2EBench.
A Reinforcement Learning Approach to Generating Test Cases for Web Applications,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.SE 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
VISCA: Inferring Component Abstractions for Automated End-to-End Testing
VISCA converts webpages into a semantic component abstraction and uses it as context for LLM-generated end-to-end tests, reporting 92% average feature coverage on E2EBench.