Prompt-engineered LLMs produced partially consistent BDI-II depression assessments in a no-ground-truth pilot, while a temporal-attention LightGBM ranked well at one writing but poorly overall in eRisk Task 2.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
DS@GT at eRisk 2025: From prompts to predictions, benchmarking early depression detection with conversational agent based assessments and temporal attention models
Prompt-engineered LLMs produced partially consistent BDI-II depression assessments in a no-ground-truth pilot, while a temporal-attention LightGBM ranked well at one writing but poorly overall in eRisk Task 2.