PhD student, Michigan State University
1 paper at NeurIPS 2025
We introduce a long-context stress test revealing that existing saliency methods fail under scale, and propose ABO—a simple, causal, and robust alternative based on attention bias optimization.