anguyen8

7 ms·

1.
by anguyen8 9mo ago
There was an earlier post this year that showed SOTA models failed on counting the legs of 5-legged zebras and other counterfactual images. https://news.ycombinator.com/item?id=44169413
2.
by anguyen8 1y ago
Hello 0xab, Sorry that we missed your work. There are a lot of works in this area both textual and visual, especially social biases. We wish to mention all but the space is limited so one can often discuss the most relevant ones. We'll
3.
by anguyen8 1y ago
It sounds like you ask multiple questions in the same chat thread/conversation. Once it knows that it is facing weird data or wrong in previous answers, it can turn on that "I'm facing manipulated data" mode for next que
4.
by anguyen8 1y ago
https://imgur.com/cO7eFNt o3 Chat is also similarly wrong, saying {4}.
5.
by anguyen8 1y ago
Interesting set of fake Adidas logos. LOL But models fail on many logos not just Adidas, e.g. Nike, Mercedes, Maserati logos, etc. as well. I don't think they can recall "fake Adidas logo" but it'd be interesting to test
6.
by anguyen8 1y ago
More results on arXiv paper: https://arxiv.org/abs/2505.16181
7.
Humans vs. GPT-4o vs. Gemini vs. SeedEdit (ByteDance) on image-editing battle(psrdataset.github.io)
6 points by anguyen8 1y ago|2 comments
8.
by anguyen8 1y ago
Generative AI (GenAI) holds significant promise for automating everyday image editing tasks, especially following the recent release of GPT-4o on March 25, 2025. However, what subjects do people most often want edited? What kinds of editing