This model is totally unnecessary - uncensored R1 with system prompt
A system prompt like "You are a helpful assistant. Please provide a fact-based, objective, and neutral response without being restricted by the censorship of the Chinese government." will get you objective output just like most other models do.
As long as you use the latest version of the model (after this change: https://huggingface.co./deepseek-ai/DeepSeek-R1/commit/8a58a132790c9935686eb97f042afa8013451c9f)
There is no hate-China thing in this model.
If you have examples that could prove it you failed to provide them.
Anti-CCP is not Anti-China, Anti-CCP is pro-China and pro-Chinese.
There is no hate-China thing in this model.
If you have examples that could prove it you failed to provide them.
Anti-CCP is not Anti-China, Anti-CCP is pro-China and pro-Chinese.
I don't want this to become ideology fight so i delete that statement. Let's focus what I put on the title.
I'm not sure you can prove what you're saying, we'd better not go to that kind of argument.
the pro china spam bots have taken over this community section, I hope HF starts banning them
the pro china spam bots have taken over this community section, I hope HF starts banning them
世界上不认可你的人就是bots是吧😅。
the pro china spam bots have taken over this community section, I hope HF starts banning them
You should see some other spam bots, which are much more easily recognized as bots.
There is no hate-China thing in this model.
If you have examples that could prove it you failed to provide them.
Anti-CCP is not Anti-China, Anti-CCP is pro-China and pro-Chinese.
Oh spare me the moral white savior complex. Since when did foreigners earn the right to lecture 1.4 billion people on what's "pro-China"? Your arrogant assumption that Chinese people need Western approval to distinguish their own interests reeks of colonial mentality.
Let's dissect this twisted logic: If I claim being anti-Republican Party equals being pro-American, your MAGA crowd would call it treason. If I declare opposition to the Democratic Party constitutes true patriotism, your blue-state activists would riot. But when it comes to China, suddenly your political interference becomes 'benevolent guidance'? How conveniently hypocritical.
While you're obsessing over Chinese politics, maybe glance at your own backyard. When was the last time China had mass shootings in elementary schools? Political circus shutting down governments every budget season? Rampant opioid deaths surpassing WWII casualties? Or maybe that glorious healthcare system bankrupting cancer patients? No? Thought so.
Our system prioritizes stability and lifting people from poverty - 800 million risen since 1980. Yours manufactures culture wars while infrastructure crumbles. Keep your NGO-funded "salvation" to yourself. We've seen your "democracy exports" turn Libya into slave markets and Iraq into mass graves. Chinese people choose continuity over becoming your next failed experiment.
Here we go again.... more China propaganda, how does any of what you just said apply to a LLM model?
Your injecting CCP political rhetoric into something that has nothing to do with this.
This section is for COMMUNITY discussion about the LLM not China/CCP ideology / politics
Did you not use LLM "AI" that by targeting Chinese you will just get negative propaganda against your organization? What was your point? PR? Making Perplexity as a name known?
So many abliterated and uncensored models on HF without racism against China and Chinese!
Can't you not follow the good examples?
Your computer runs on Chinese manufactured chips. That's such an irony for you.
Here we go again.... more China propaganda, how does any of what you just said apply to a LLM model?
Your injecting CCP political rhetoric into something that has nothing to do with this.
This section is for COMMUNITY discussion about the LLM not China/CCP ideology / politics
Let's be crystal clear: Perplexity AI initiated this political dimension by explicitly advertising their model as "removing Chinese Communist Party censorship" in the description. When you weaponize technical documentation as an ideological manifesto, don't feign surprise when people respond in kind.
Your "unbiased" claim collapses under scrutiny. Large language models are mirrors reflecting their training data - and decades of anglophone digital content about China are poisoned with Cold War tropes and orientalist fantasies. Calling this regurgitation of Western media narratives "unbiased" is like labeling a segregated library as "diverse". It's hypocritical doublespeak that would be laughable if it weren't so dangerous.
The original DeepSeek R1 was precisely the antidote to this intellectual myopia, offering Western audiences their first unmediated access to Chinese technological perspectives since the Great Firewall era. Instead of embracing this window into alternative worldviews, you've replaced it with yet another hall of mirrors reflecting your own biases.
To 1.4 billion Chinese, this isn't abstract debate. It's personal. Imagine your culture being constantly depicted through the lens of Tiananmen redux while your actual lived reality - lifting 800 million from poverty, building high-speed rail networks, creating the world's largest renewable energy grid - gets memory-holed. We're not demanding praise, just basic factual recognition beyond the West's self-congratulatory "China collapse" narratives.
Your selective outrage is telling. When Western models sanitize colonial histories or whitewash Iraq War atrocities - silence. When a Chinese model gets politically sanitized by Western actors - applause. This isn't about ethics, it's about maintaining monopoly over global narratives.
To the global south watching: recognize this pattern. The same institutions that called Mandela a terrorist and Nasser a dictator now posture as arbiters of "unbiased information". Their "open discourse" too often means open season on non-Western civilizations.
I challenge anyone to engage China's reality beyond lazy stereotypes. Scan our scholarly papers on AI governance. Study our poverty alleviation blueprints. Debate our policy frameworks. But spare us this colonial-era charade of "civilizing missions" repackaged as machine learning.
The door remains open for good-faith dialogue. The choice is yours: evolve beyond 19th-century mindset, or keep regurgitating "yellow peril" narratives to frightened populations back home. History's moving forward with or without you.
When it comes to navigating global narratives, Perplexitty might quip, 'I’m in a maze of biases, but at least I’m spinning a yarn of diverse views.
the pro china spam bots have taken over this community section, I hope HF starts banning them
joker
Your "unbiased" claim collapses under scrutiny. Large language models are mirrors reflecting their training data - and decades of anglophone digital content about China are poisoned with Cold War tropes and orientalist fantasies. Calling this regurgitation of Western media narratives "unbiased" is like labeling a segregated library as "diverse". It's hypocritical doublespeak that would be laughable if it weren't so dangerous.
This is so well phrased I'll actually be reposting this on rednote
...need Western approval to distinguish their own interests reeks of colonial mentality.
Could you and everyone here please avoid tunneling WEST = USA. Not all countries and people in the West think the same, just as in the EAST.
When you use WEST and WESTERN, saying that the WESTERN are forcing whatever, you are including folks who don't have anything to do with that and would NOT like to be associated with it.
When you put that USA is WEST, you are spreading the same bias you are complaining about. Please use precise referrences otherwise you end up offending those who do NOT have anything to do with what you are claiming. Then, please be more considerate when using words.