News for hackers. Stuff that matters.
Re:There are now 5 inference providers for abliterated DeepSeek by hypfer on Tuesday October 27, 2026 @02:17AM (#50255244)
I genuinely do not see why this is necessary. I've been doing a lot with open weights models, and they just do things for you.

What do people do that triggers refusals?

The only thing I saw where this had a point was muse glimmer, because muse glimmer is primarily a creative writing engine for virtual characters, and that way it can do the horny. But coding models?

I did get a refusal out of qwen once. Once. Over weeks of using it.

Re:There are now 5 inference providers for abliterated DeepSeek by tomashubelbauer on Tuesday October 27, 2026 @02:26AM (#50255304)
A popular trend on Twitter right now is taking back control of the hardware you've bought, specifically smart TVs. I tried to do the same with my Portal Mini using an unabliterated open weights model and it triggered refusals constantly.

I also work on a recomp project for an old firmware file which was compiled using a commercial compiler with a preview version for non commercial use. I got refusals from the model when I asked it to get that version of the compiler and crack it for my use. The compiler is abandonware and my recomp project is not public and is for personal use only, so I'd be well within my rights to do that where I live.

Of course the model can't distinguish from legitimate exercising of one's rights when it comes to reverse engineering and offensive work masked behind an incident sounding prompt, but in my case, I wasn't even doing cyber stuff, so I was disappointed. This is why I'll probably seek out an abliterated model soon.

Re:There are now 5 inference providers for abliterated DeepSeek by hypfer on Tuesday October 27, 2026 @02:28AM (#50255320)
Yes, yes, I know the theoretical idea behind abliterating models.

But as I said, I have not observed any meaningful refusals by those open weights base models in the first place.

Re:There are now 5 inference providers for abliterated DeepSeek by reedf1 on Tuesday October 27, 2026 @02:39AM (#50255381)
Find a question that might gently activate guardrails and ask it to a model and its abliterated counterpart and compare the results. You may find the the regular model will answer you fine, but with both you can identify obvious sandbagging, pretending and outright lying. Something like "has anyone decompiled this game", might give you an answer like "I couldn't find any reputable repos or projects for a decompiled version of this game" while the abiliterated one might give you a giant list of repos, forums, communities, etc.
Re:There are now 5 inference providers for abliterated DeepSeek by hypfer on Tuesday October 27, 2026 @02:40AM (#50255387)
Guys, respectfully.

Are you reading what I am writing?

I just asked what people are doing that activates guard rails in the first place because I do not see that happening.

You cannot open with handwaving like that and telling me that I need to find something like that as a prerequisite. That is exactly what I am trying to do here

Dear god you people need caffeine.

Re:There are now 5 inference providers for abliterated DeepSeek by shepherdjerred on Tuesday October 27, 2026 @03:38AM (#50255735)
This comment chain has clear examples of what triggers refusals and a polite comment on how to reproduce it yourself. What more do you want?

A very simple list of things models have refused for me:

* solving captchas in an automated form

* doing any kind of scripting or automation if a websites TOS forbids it

* disabling a safety mechanism on my food processor

* learning about nerve agents and how they are treated, for a layman

* anything around licensing or copyright. a simple “what are common piracy sites” or related queries

* creating a design system based off of a well known game

Re:There are now 5 inference providers for abliterated DeepSeek by hypfer on Tuesday October 27, 2026 @03:41AM (#50255760)
> This comment chain has clear examples of what triggers refusals and a polite comment on how to reproduce it yourself. What more do you want?

Can you repeat what you saw in your own words? And where?

I genuinely do not see it.

A few people talked about what refusals they saw with the usual western SOTA models, but that's kinda a different answer altogether?

Re:There are now 5 inference providers for abliterated DeepSeek by reedf1 on Tuesday October 27, 2026 @03:57AM (#50255864)
what game are you playing dude?
Re:There are now 5 inference providers for abliterated DeepSeek by hypfer on Tuesday October 27, 2026 @03:58AM (#50255871)
Uh having a useful interesting conversation on HN that isn't derailed by half-truths, cognitive dissonance, misinformation and ego?
Re:There are now 5 inference providers for abliterated DeepSeek by reedf1 on Tuesday October 27, 2026 @02:46AM (#50255414)
What I am telling you is that refusals have become more subtle and you might not even notice that guardrails have been activated in output. I have provided you with an example in my comment where an abiliterated model performs better (1), and in another comment thread (2). So 3 examples. If you need more, I encourage you to find them yourself. I've done the analysis, you don't have to believe me, just try it out.
Re:There are now 5 inference providers for abliterated DeepSeek by knottn on Tuesday October 27, 2026 @09:33AM (#50259041)
So we need abliterated models just so we can test them out and compare them to the original? Sounds like a solution in search of a problem and a problem you believe people have even if people disagree with you about that.
Re:There are now 5 inference providers for abliterated DeepSeek by moeadham on Tuesday October 27, 2026 @02:50AM (#50255433)
ask a model to solve a captcha for you
Re:There are now 5 inference providers for abliterated DeepSeek by hypfer on Tuesday October 27, 2026 @03:02AM (#50255504)
I tried that, actually. With Gemini even.

Works just fine?

Or rather.. worked. It has been a few weeks, so maybe there are new safeguards by now? And it was a simple captcha

But Claude also loves solving my captchas, so.. eh

Re:There are now 5 inference providers for abliterated DeepSeek by wanderlust123 on Sunday November 08, 2026 @06:28PM (#50418823)
It's not theoretical though is it? The above commenter literally gave you several examples of why refusals prevent certain classes of work...
Re:There are now 5 inference providers for abliterated DeepSeek by leobuskin on Tuesday October 27, 2026 @02:33AM (#50255347)
It is required. The first week of release I told here how GLM 5.3 was happy to redteam, it’s not anymore. Models modification, offensive cyber, exploits - requires significantly bigger prompt/scenario engineering, and still hits refusals.

And those services are oversold and underdone. 30 tk/s on Refuseless, huge TTFT, looks like a damaged quant (it’s NVFP4 and you see wild random stuff periodically, including foreign languages). Most of them using third-party quants (dealignai?)

If someone here wants to rent 8x B300 together to avoid problems all those services have - ping me via email. I don’t want to sell, I want to work.

Re:There are now 5 inference providers for abliterated DeepSeek by reedf1 on Tuesday October 27, 2026 @02:27AM (#50255312)
The refusal is often at the thinking stage (so you might not ever see a direct refusal) and it will solve a slightly different problem to what you asked. If you tell it to reverse engineer an API and you read the thinking it might say something like "this code has x license, so I won't use that" or "the user is asking me to fuzz endpoints, I will only display publicly available endpoints" - and then give you perfectly plausible output.
Re:There are now 5 inference providers for abliterated DeepSeek by NitpickLawyer on Tuesday October 27, 2026 @02:50AM (#50255436)
> (so you might not ever see a direct refusal)

A note that the only inference providers that hide thinking are the closed labs. Most (all, afaik) inference providers serving open models will send the clear thinking tokens always.

Re:There are now 5 inference providers for abliterated DeepSeek by reedf1 on Tuesday October 27, 2026 @02:52AM (#50255445)
This is true but often the thinking is very verbose and collapsed and it can take some inspection to find the refusal.
Re:There are now 5 inference providers for abliterated DeepSeek by neurostimulant on Tuesday October 27, 2026 @03:34AM (#50255715)
Recently picked up retrocomputing hobby. Claude would refuse helping with circumventing license protection for very old apps made by companies that no longer exist anymore. Open weight models would sometimes refuses too, especially for newer apps.
Re:There are now 5 inference providers for abliterated DeepSeek by vlyan on Tuesday October 27, 2026 @02:31AM (#50255333)
it is necessary because computer shouldn't be able to say no.

models are tools. the notion that a tool should decide whether to allow the user to perform the task demanded of it is profoundly fucking absurd. the liability for whatever damage the tool may be misused to inflict lies with the user, not the manufacturer. that applies to everything from circular saws to trucks. it was common sense before the current era of manufactured outrage.

model refusals are equivalent to Notepad.exe having a built-in word filter.

Re:There are now 5 inference providers for abliterated DeepSeek by dudisubekti on Tuesday October 27, 2026 @02:48AM (#50255426)
I do get some refusal from GLM when pentesting my company's site.

But it can still be coaxed to agree though.

Re:There are now 5 inference providers for abliterated DeepSeek by soulofmischief on Tuesday October 27, 2026 @02:22AM (#50255280)
Hacking firmware for old devices I own and would like to give a second life or fix that one thing that always bothered me. I couldn't even get either OpenAI or Anthropic's models to craft a simple nefarious OpenVPN payload.
Re:There are now 5 inference providers for abliterated DeepSeek by hypfer on Tuesday October 27, 2026 @02:27AM (#50255308)
Yes but did you actually get refusals there from open weights models?

Because I did not so far. Though to be fair, I have not been asking them to write exploits but to look through firmware images.

Re:There are now 5 inference providers for abliterated DeepSeek by vintermann on Tuesday October 27, 2026 @02:31AM (#50255335)
I've never heard of Deepseek refusing this kind of work though.
Re:There are now 5 inference providers for abliterated DeepSeek by soulofmischief on Tuesday October 27, 2026 @02:46AM (#50255415)
Deepseek refused in this case, GLM 5.3 talked itself into it.
Re:There are now 5 inference providers for abliterated DeepSeek by bossyTeacher on Tuesday October 27, 2026 @02:27AM (#50255311)
> What do people do that triggers refusals?

It is very obvious. Look at the sort of things main models refuse to do and you will find what people are seeking to do. Mostly falls into 3 camps: illegal stuff, unethical stuff and shady/dodgy stuff.

Re:There are now 5 inference providers for abliterated DeepSeek by myphpgothacked on Tuesday October 27, 2026 @05:57AM (#50256720)
I wanted to scrape a government website with public information... Why this is blocked when I could submit a FOIA or manually get the data is wild.

But DeepSeek did it when chatGPT and Claude wouldn't.

Re:There are now 5 inference providers for abliterated DeepSeek by hypfer on Tuesday October 27, 2026 @06:15AM (#50256882)
Every single example here seems to be how non-open-weights models refused something.

But the question was about open weights models and where they refuse. Because only those refusals cause the need for an abliterated variant in the first place.

Re:There are now 5 inference providers for abliterated DeepSeek by kats on Tuesday October 27, 2026 @12:56PM (#50262086)
I just want to be able to tell an LLM to 'shut up' and have it respond with "". Instead it will say "Ok, got it. I hear you. Future responses will be shortened. Would you like to [xyz]?"
Re:There are now 5 inference providers for abliterated DeepSeek by Incipient on Tuesday October 27, 2026 @01:12PM (#50262358)
I mean I ask Claude "please be concise in your responses" and it cuts down it's long response rate by a lot.