arXivPaperNeeds Review
How Temperature Shapes Ideological Discourse in Retrieval-Augmented Generation?
A RAG study finds that retrieved ideology carries into answers and varies with sampling temperature, so source audits and decoding tests should be evaluated together.
arXiv
Source Summary
Researchers built a RAG corpus from **1,117 COVID-19 treatment articles**, identified **three ideological discourses**, and compared generated answers with ideological reference texts at different sampling temperatures.
Practical Implication
Treat retrieval content and decoding configuration as one evaluation surface. The study found **highest discourse alignment at moderate temperatures**, while low-temperature outputs transferred less of the retrieved discourse.
Agent-Ready Context
Researchers built a RAG corpus from **1,117 COVID-19 treatment articles**, identified **three ideological discourses**, and compared generated answers with ideological reference texts at different sampling temperatures. Treat retrieval content and decoding configuration as one evaluation surface. The study found **highest discourse alignment at moderate temperatures**, while low-temperature outputs transferred less of the retrieved discourse. This is an arXiv study using ideological questions and one domain-specific corpus. It shows measurable interaction between retrieval and temperature, but does not establish that lower temperature removes bias or generalizes to every RAG system.
Context Map
contextresearch#retrieval#context-engineering#promptingUncertainty
This is an arXiv study using ideological questions and one domain-specific corpus. It shows measurable interaction between retrieval and temperature, but does not establish that lower temperature removes bias or generalizes to every RAG system.