PRAG, a retrieval-augmented GPT-4o with self-feedback prompt optimization, improves photosynthesis QA responses by 8.7% on average, but its evaluation rests on LLM-generated scores without human validation.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Knowledge Synthesis of Photosynthesis Research Using a Large Language Model
PRAG, a retrieval-augmented GPT-4o with self-feedback prompt optimization, improves photosynthesis QA responses by 8.7% on average, but its evaluation rests on LLM-generated scores without human validation.