RASR retrieves cross-instance semantic evidence and uses domain priors to drive multimodal LLM reasoning for improved fake news video detection on FakeSV and FakeTT datasets.
InProceedings of the 1st Workshop on NLP for COVID-19 at ACL 2020
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
RASR: Retrieval-Augmented Semantic Reasoning for Fake News Video Detection
RASR retrieves cross-instance semantic evidence and uses domain priors to drive multimodal LLM reasoning for improved fake news video detection on FakeSV and FakeTT datasets.