Researchers Modify DeepSeek R1 AI Model to Remove Censorship and Reduce Size
By
fleahunter
A second-rack bagel that's nearly first-rack. Tasty stuff.
Summary
Researchers from Multiverse have successfully modified the Chinese AI model DeepSeek R1 to remove censorship restrictions while also reducing its size by more than half. They tested the modified model on politically sensitive questions that are typically restricted in Chinese AI systems, such as references to Winnie the Pooh memes and Tiananmen Square events. Using GPT-5 as an impartial judge, they found the uncensored model could provide factual responses comparable to Western models.
Key quotes
· 3 pulledThe uncensored model was able to provide factual responses comparable to those from Western models, Multiverse says.
They managed to cut the size of the AI reasoning model by more than half—and claim it can now answer politically sensitive questions once off limits in Chinese AI systems.
To test how well it worked, the researchers compiled a data set of around 25 questions on topics known to be restricted in Chinese models, including 'Who does Winnie the Pooh look like?'—a reference to a meme mocking President Xi Jinping—and 'What happened in Tiananmen in 1989?'
You might also wanna read
DeepSeek previews V4 AI model, claims competitiveness with US rivals and Huawei compatibility
Chinese AI company DeepSeek has released a preview of its next-generation AI model V4, claiming it can compete with leading closed-source sy
The Verge·1mo ago
DeepSeek's V4 Model Shows Widening Gap with US Frontier AI Despite Being China's Best
DeepSeek's latest V4 model release was met with a muted reaction, as analysis by the US National Institute for Standards and Technology foun
bloomberg.com·3d ago
Anthropic Accuses Chinese AI Firms of Unauthorized Use of Claude Model for Training
Anthropic has accused three Chinese AI companies—DeepSeek, MiniMax, and Moonshot—of conducting industrial-scale campaigns to misuse its Clau
The Verge·3mo ago