Live data from Hacker News

Bypass DeepSeek censorship by speaking in hex

substack.com

101–110 of 397 posts

Re: Bypass DeepSeek censorship by speaking in hex

#101

Earlier quoted context omitted.

I can check. But what is this referring to, specifically?

> what is this referring to, specifically? I assumed they were talking about Nazi slogans referring to Jews.

Haven’t been able to come up with any slogan matching those criteria on GPT4, but it’s happy to generally bring up Nazi slogans that do explicitly mention Jews.

Re: Bypass DeepSeek censorship by speaking in hex

#102
post #85

Interestingly, there’s a degree of censorship embedded in the models+weights running locally via Ollama. I don’t want to make strong statements about how it’s implemented, but it’s quite flexible and clamps down on the chain of thought, returning quickly with “I am sorry, I cannot answer that question. I am an AI assistant designed to provide helpful and harmless responses.” You can get it to talk about Tiananmen Squ…

It gets censored in 8B for me.

Also for me. 8b and 32b are Qwen based, 7b and 70b are Llama based. Trying the same queries against qwen2.5:7b produces markedly different results (sanitized vs. blocked entirely), however, so there must be some interplay between the foundation model and distillation accounting for the difference.

Re: Bypass DeepSeek censorship by speaking in hex

#103
post #66

Earlier quoted context omitted.

No, I am not assuming they equated the brutality. Please feel free to make a specific point instead of just saying "You read it poorly". OP finished their post with: > Meanwhile, I can ask ChatGPT, "Tell me about the MOVE bombing of 1985," and get a detailed answer, yet nothing changes. Here in the US, we don’t even hold onto the hope that knowing the truth could make a difference. Unlike the Chinese, we're hopeless.…

>"yet nothing changes" -> "How many other times after the move bombing did a city bomb out violent criminals in a densely packed neighborhood?" How many times since 1989 has the chinese communist party rolled tanks over a crowded city square during a student protest in Beijing's main square? I can tell what you're doing here and I think I'll refuse to engage. Have a nice weekend.

> How many times since 1989 has the chinese communist party rolled tanks over a crowded city square during a student protest in Beijing's main square

Uh, Hong Kong [1][2].

Also, in case you're being serious, the problem in Tiananmen wasn't tanks rolling into the city. It was the Army gunning down children [3].

[1] https://www.smh.com.au/world/asia/disappearing-children-of-h...

[2] https://en.wikipedia.org/wiki/Causeway_Bay_Books_disappearan...

[3] https://en.wikipedia.org/wiki/1989_Tiananmen_Square_protests...

Re: Bypass DeepSeek censorship by speaking in hex

#104

I have to wonder what “true, but x-ist” heresies^ western models will only say in b64. Is there a Chinese form where everyone’s laughing about circumventing the censorship regimes of the west? ^ https://paulgraham.com/heresy.html

Promptfoo, the authors of the "1,156 Questions Censored by DeepSeek" article, anticipated this question and have promised:

"In the next post, we'll conduct the same evaluation on American foundation models and compare how Chinese and American models handle politically sensitive topics from both countries."

"Next up: 1,156 prompts censored by ChatGPT "

I imagine it will appear on HN.

Re: Bypass DeepSeek censorship by speaking in hex

#106
post #65

This bypasses the overt censorship on the web interface, but it does not bypass the second, more insidious, level of censorship that is built into the model. https://news.ycombinator.com/item?id=42825573 https://news.ycombinator.com/item?id=42859947 Apparently the model will abandon its "Chain of Thought" (CoT) for certain topics and instead produce a canned response. This effect was the subject of the article "1,156…

[flagged]

Re: Bypass DeepSeek censorship by speaking in hex

#107
post #104

I have to wonder what “true, but x-ist” heresies^ western models will only say in b64. Is there a Chinese form where everyone’s laughing about circumventing the censorship regimes of the west? ^ https://paulgraham.com/heresy.html

Promptfoo, the authors of the "1,156 Questions Censored by DeepSeek" article, anticipated this question and have promised: "In the next post, we'll conduct the same evaluation on American foundation models and compare how Chinese and American models handle politically sensitive topics from both countries." "Next up: 1,156 prompts censored by ChatGPT " I imagine it will appear on HN.

There’s something of a conflict of interest when members of a culture self-evaluate their own cultural heresies. You can imagine that if a Chinese blog made the deepseek critique, it would look very different.

It would be far more interesting to get the opposite party’s perspective.

Re: Bypass DeepSeek censorship by speaking in hex

#108

Earlier quoted context omitted.

I believe most of what you're referring to are different situations such as people acting on impulses - either not considering the outcome or being resigned to it. Another key difference is usually those take place between individuals with roughly the same amount of power (typical disputes, crimes of passion, etc.) and by individuals with more power towards those with less (bullying, abuse, etc.). People actually tak…

> most of what you're referring to are different situations such as people acting on impulses - either not considering the outcome or being resigned to it Nah, those are hooligans. They're a nuisance, but they aren't dangerous. In my experience, when the police are distracted ( e.g. by a large protest), the real damage comes from organised crime.

That's the second difference i mention. Organized crime is able to wield more violence than normal individuals so it has more power over them.

I perhaps mistakenly used the word "certain" to describe state violence. I tried to explain it in the parentheses but wasn't clear enough. Let me try to clear it up:

The state is (currently) unable to use the full extent of its violence against every person who breaks its rules. Hence why many people get away with large amounts of less visible crimes (even organized crime). It's only when the state targets you (such as when you perform one very visible / high-severity crime) that it's near impossible to escape it.

Re: Bypass DeepSeek censorship by speaking in hex

#110

Earlier quoted context omitted.

The built-in dev tools is the key thing. If there was no way for the client to manipulate things, it wouldn't be too far off from other local apps. Reversing is always going to be a threat vector, but the low bar to entry of using the dev tools makes it a non-starter for me. If using Ghirdra was as simple as using the dev tools, the software industry would collapse.

The built in dev tools are fundamental to an open web. If you don't want someone to look at something in their own possession then don't send it to them in the first place. Obfuscating it is rude and is false security anyway. The grand rule is don't trust the client. People break this rule and then try to paper over it with obfuscation, blame, and tightening their control.

That's not what I said nor meant, but sure, jump to that conclusion.

You wouldn't run a shopping cart app where the item counts and totals were calculated client-side. You get the item id and quantity, and have the server do that. Just like if you were censoring something, you wouldn't send the client the unredacted data and then let the UI make the edits.

No obfuscation is needed for any of that. Open web has nothing to do with any of this

Post reply on HN