Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
Tinted Frames: Question Framing Blinds Vision-Language Models
(
davidhalladay.github.io
)
4 points
by
ritwikgupta
6 months ago
|
hide
|
past
|
favorite
|
1 comment
ritwikgupta
6 months 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.
Consider applying for YC's Winter 2027 batch!
Applications
are open till November 2.
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search:
