TY  - RPRT
TI  - Ablate-to-Validate: Are Vision-Language Models Really Using Continuous Thought Tokens?
AU  - Tianyi Zhang
AU  - Mahtab Bigverdi
AU  - Ranjay Krishna
PY  - 2026
UR  - https://arxiv.org/abs/2605.21642
ID  - 2605.21642
ER  -