Nightshade: Prompt-Specific Poisoning Attacks on Text-to-Image Generative Models · HackerPapers discussions

Nightshade: Prompt-Specific Poisoning Attacks on Text-to-Image Generative Models — mentioned by torginus

Sun, 04 Jan 2026 16:24:47 GMT

It is kind of unfortunate how people don't actually read the paper but only run with the conclusions, speculating whether this would or would not work. Here's the paper in question: https://arxiv.org/abs/2310.13828 My two cents is that in its current implementation the compromised images can be easily detected, and possibly even 'de-poisoned'. The attack works by targeting an underrepresented concept (let's say 1% of images contain dogs, so 'dog' is a good concept to attack). They poison the concept of 'dog' with the concept of 'cat' by blending (in latent space) an archetypical image of…

Paper and discussion history. A mention is not an endorsement.

Nightshade: Prompt-Specific Poisoning Attacks on Text-to-Image Generative Models — HN submission

Tue, 06 Feb 2024 02:43:34 GMT

Prompt-Specific Poisoning Attacks on Text-to-Image Generative Models (2023)

3 points; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Nightshade: Prompt-Specific Poisoning Attacks on Text-to-Image Generative Models — mentioned by brucethemoose2

Sat, 20 Jan 2024 18:41:37 GMT

What the article doesn't illustrate is that it destroys fine detail in the image, even in the thumbnails of the reference paper: https://arxiv.org/pdf/2310.13828.pdf Also... Maybe I am naive, but it seems rather trivial to work around with a quick prefilter? I don't know if tradition denoising would be enough, but worst case you could run img2img diffusion. reply

Paper and discussion history. A mention is not an endorsement.

Nightshade: Prompt-Specific Poisoning Attacks on Text-to-Image Generative Models — mentioned by brucethemoose2

Sat, 20 Jan 2024 18:41:03 GMT

Paper and discussion history. A mention is not an endorsement.

Nightshade: Prompt-Specific Poisoning Attacks on Text-to-Image Generative Models — mentioned by ink404

Fri, 19 Jan 2024 17:48:58 GMT

Paper is here: https://arxiv.org/abs/2310.13828

Paper and discussion history. A mention is not an endorsement.

Nightshade: Prompt-Specific Poisoning Attacks on Text-to-Image Generative Models — HN submission

Thu, 04 Jan 2024 03:46:35 GMT

Prompt-Specific Poisoning Attacks on Text-to-Image Generative Models

1 point; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Nightshade: Prompt-Specific Poisoning Attacks on Text-to-Image Generative Models — HN submission

Thu, 14 Dec 2023 14:40:42 GMT

Prompt-Specific Poisoning Attacks on Text-to-Image Generative Models

1 point; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Nightshade: Prompt-Specific Poisoning Attacks on Text-to-Image Generative Models — HN submission

Fri, 27 Oct 2023 12:30:21 GMT

Prompt-Specific Poisoning Attacks on Text-to-Image Generative Models

1 point; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Nightshade: Prompt-Specific Poisoning Attacks on Text-to-Image Generative Models — HN submission

Wed, 25 Oct 2023 21:37:28 GMT

Prompt-Specific Poisoning Attacks on Text-to-Image Generative Models

2 points; 1 discussion reply on HN. Paper and source history. A submission is not an endorsement.

Nightshade: Prompt-Specific Poisoning Attacks on Text-to-Image Generative Models — mentioned by xyzal

Wed, 25 Oct 2023 14:46:24 GMT

Paper> https://arxiv.org/abs/2310.13828

Paper and discussion history. A mention is not an endorsement.

Nightshade: Prompt-Specific Poisoning Attacks on Text-to-Image Generative Models — HN submission

Tue, 24 Oct 2023 14:47:48 GMT

Prompt-Specific Poisoning Attacks on Text-to-Image Generative Models

2 points; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.