Scaling Trends for Data Poisoning in LLMs · HackerPapers discussions

Scaling Trends for Data Poisoning in LLMs — mentioned by meander_water

Fri, 10 Oct 2025 02:02:24 GMT

I don't think this is a bombshell finding. Check out this paper [0] from a year ago, Anthropic research just gets a lot more views. > Our experiments reveal that larger LLMs are significantly more susceptible to data poisoning, learning harmful behaviors from even minimal exposure to harmful data more quickly than smaller models. [0] https://arxiv.org/html/2408.02946v4

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

Scaling Trends for Data Poisoning in LLMs — mentioned by pera

Thu, 16 Jan 2025 15:12:59 GMT

Does anyone know if there is anything like Nepenthes but that implements data poisoning attacks like https://arxiv.org/abs/2408.02946

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

Scaling Trends for Data Poisoning in LLMs — HN submission

Mon, 13 Jan 2025 16:33:51 GMT

Data Poisoning in LLMs: Jailbreak-Tuning and Scaling Laws

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