ChatGPT can produce synthetic system requirement specifications that 62 percent of experts rate as realistic, though the outputs contain contradictions and deficiencies.
Can you trust llm judgments? reliability of llm-as-a-judge, 2025
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.SE 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Can ChatGPT Generate Realistic Synthetic System Requirement Specifications? Results of a Case Study
ChatGPT can produce synthetic system requirement specifications that 62 percent of experts rate as realistic, though the outputs contain contradictions and deficiencies.