Tinted Frames: Question Framing Blinds Vision-Language Models

Tinted Frames: Question Framing Blinds Vision-Language Models(davidhalladay.github.io)

4 points by ritwikgupta 166 days ago | 1 comment

ritwikgupta 166 days ago |

VLMs are selectively blind — they decide how much to look at an image based on question framing (open-ended vs Yes/No vs MCQ), even when the same visual reasoning is required.

ritwikgupta 166 days ago |

VLMs are selectively blind — they decide how much to look at an image based on question framing (open-ended vs Yes/No vs MCQ), even when the same visual reasoning is required.