Persistent Pre-Training Poisoning of LLMs · HackerPapers discussions

Persistent Pre-Training Poisoning of LLMs — mentioned by Terr_

Sun, 01 Nov 2026 07:47:59 GMT

This is the continuation of a problem that's been known-about for a few years. [0][1] I suppose the news is that yes, the issue remains in modern models. LLMs driving decisions are going to be like asbestos: People will ignore the risk for a while to make a short-ter profit, eventually it will become too hard to ignore, and then it'll be expensive to remove and replace it with a technology that doesn't have the same dangers. [0] 3 years ago, "Universal Jailbreak Backdoors from Poisoned Human Feedback" - https://arxiv.org/abs/2311.14455 [1] 2 years ago, "Persistent Pre-Training Poisoning of…

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

Persistent Pre-Training Poisoning of LLMs — mentioned by Tiberium

Mon, 24 Aug 2026 07:31:40 GMT

There is quite old research on this: - https://arxiv.org/abs/2311.14455 - https://arxiv.org/abs/2401.05566 - https://arxiv.org/abs/2410.13722

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

Persistent Pre-Training Poisoning of LLMs — mentioned by hazrmard

Tue, 18 Aug 2026 15:56:24 GMT

I read this paper (https://arxiv.org/pdf/2410.13722) on pre-training poisoning of LLMs. The authors demonstrated that poisoning 0.1% of the pretraining dataset can measurably manipulate beliefs of the model. What I read about is a more sophisticated supply chain attack. But these days, search previews use retrieval-augmented generation to parse websites to generate answers on the fly. In that case you don't need to poison the supply chain; good SEO is enough.

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

Persistent Pre-Training Poisoning of LLMs — mentioned by nyrikki

Mon, 23 Dec 2024 21:06:02 GMT

This study shows that controlling 0.1% may be enough. https://arxiv.org/abs/2410.13722v1 I have noticed some popular copied but incorrect leetcode examples leaking into the dataset. I suspect it depends on domain specificity, but that seems within the ability of an SEO spammer or decentralized group of individuals.

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

Persistent Pre-Training Poisoning of LLMs — HN submission

Mon, 21 Oct 2024 20:16:59 GMT

Persistent Pre-Training Poisoning of LLMs

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