Empirical test on real news histories shows unconstrained LLM reranking amplifies conspiratorial content exposure while prompt constraints can increase ideological diversity with limited relevance cost.
Title resolution pending
4 Pith papers cite this work. Polarity classification is still indexing.
4
Pith papers citing it
verdicts
UNVERDICTED 4representative citing papers
Reddit analysis finds MAHA users show strong cross-theme belief bundling and network coherence unlike anti-MAHA users, with pandemic-era shifts from anti-fluoride/mask to anti-vaccine to broader anti-science engagement.
Republican posts are more toxic than Democratic ones, but Democratic content draws more toxic replies due to higher volume of Republican cross-partisan engagement.
Social media research yields inconclusive causal findings due to system complexity, and progress requires mechanistic explanations that integrate observational and experimental approaches while recognizing their shared limitations.
citing papers explorer
-
The Structure and Dynamics of the Online MAHA-sphere
Reddit analysis finds MAHA users show strong cross-theme belief bundling and network coherence unlike anti-MAHA users, with pandemic-era shifts from anti-fluoride/mask to anti-vaccine to broader anti-science engagement.
-
"F*** You Biden": Cross-Partisan Electoral Toxicity on X
Republican posts are more toxic than Democratic ones, but Democratic content draws more toxic replies due to higher volume of Republican cross-partisan engagement.
-
Moving towards informative and actionable social media research
Social media research yields inconclusive causal findings due to system complexity, and progress requires mechanistic explanations that integrate observational and experimental approaches while recognizing their shared limitations.