Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
1
0
0
Contradicting the assumption that MLLMs require specialized forensic fine-tuning to spot subtle image edits, simply contrasting cross-attention against a Gaussian-blurred baseline isolates genuine manipulation artifacts and cuts explanatory hallucination by up to 33.4% training-free.