TIGER: Traceable Inference with Graph-Based Evidence Routing for Mitigating Hallucinations in Multimodal Generation

In three linesTIGER is an inference-time framework to mitigate hallucinations in multimodal generation. It independently extracts an observation graph from input and a claim graph from output, then assigns risk scores to claims based on support and conflict. The model repairs high-risk claims while keeping the backbone frozen. Convergence analysis shows geometric risk reduction to an explicit asymptotic bound.

Summary generated by Claude — human-verified