2 papers
cs.CV2026
Does Video Memory Use What It Retrieves? A Causal Audit of Memory Specificity
Aditi Tiwari, Akshit Bhalla, Darshan Prasad +1
Video models increasingly use memory to preserve information over long sequences, with the assumption that gains come from retrieving and using the correct past content. Standard m…
cs.CV2026
Inference-Time Attention Steering for Vision-Language-Action Driving Models
Darshan Nagendra Prasad, Lars Ullrich, Knut Graichen
Vision-language-action (VLA) driving models couple a reasoning stage with a diffusion-based trajectory decoder, but do not give a direct way to redirect attention toward safety-cri…