Perplexity has launched the R1 1776 model, a refined version of Deepseek R1, featuring open model weights aimed at eliminating Chinese censorship. This move, articulated by co-founder Aravind Srinivas, emphasizes that the model's reasoning capabilities remain intact despite the removal of biases. With R1 1776, users can now obtain unfiltered information on sensitive topics, enhancing the quality and accuracy of responses in a broader context. This development aims to provide clarity on crucial issues, including governance and economic consequences related to Taiwan's independence.
The post-training to remove censorship has been carried out without degrading the basic reasoning ability of the model, which is crucial for the model to remain useful in all practically important tasks.
R1 1776 has post-trained to remove China's censorship and provide more unbiased, accurate answers, ensuring that users receive clear and direct responses.
Collection
[
|
...
]