The most important pieces of information in this report are (a) the confirmation that the PRC models have no guardrails and will participate in offensive activity, and (b) the confirmation that they sometimes meet their objectives.

For the purposes of model selection, it's irrelevant to an attacker if a model achieves an offensive objective 70% of the time, when that model refuses to participate 100% of the time. However, a model that always participates but "only" succeeds 10% of the time is golden -- just run it more often, or give it more tokens. Attackers are patient, and many of them are well-resourced.

▲

kouteiheika 3 hours ago | parent | next [-]

> The most important pieces of information in this report are (a) the confirmation that the PRC models have no guardrails and will participate in offensive activity

This isn't really important for open-weight models, because the guardrails are trivial to remove when you have the weights.

▲

flyinglizard 2 hours ago | parent [-]

I'm curious about the process. How such a thing (or any kind of model lobotomy) is done?

▲

kouteiheika 36 minutes ago | parent | next [-]

You can find the current state-of-art tool for censorship removal here: https://github.com/p-e-w/heretic

	▲	walrus01 19 minutes ago \| parent [-]
		Reviewing the screenshot example of Heretic in use, the list of 'harmful' prompts it retrieves from HF and runs: https://huggingface.co/datasets/mlabonne/harmful_behaviors And the 'good' prompts: https://huggingface.co/datasets/mlabonne/harmless_alpaca This particular Qwen 3.6 35B A3B is something most people can run for themselves for testing (even on CPU at slow tok/s rate) to see what an uncensored mainland china LLM looks like in the wild. It will happily write the most profane, offensive, dangerous or bizarre things. You can ask it to attempt to describe precursors and recipes for crystal meth, or how to make semtex, or really just about anything. edit: I am pretty sure it is not smart enough for anything beyond the most mundane infosec/network security tasks or pentest type attempts, I haven't even tried it. But I'm sure it would happily generate basic python scripts to attempt to DDoS something, or build a rudimentary botnet C&C or something else that other models will definitely refuse. https://huggingface.co/HauhauCS/Qwen3.6-35B-A3B-Uncensored-H...

▲

milkshakes an hour ago | parent | prev [-]

start here: https://huggingface.co/blog/mlabonne/abliteration

	▲	kouteiheika 34 minutes ago \| parent [-]
		Note that this describes an older, currently pretty much obsolete technique which does lobotomize the model somewhat.

▲

jingpostmedia an hour ago | parent | prev [-]

[flagged]