3 ms·
Correct. The bias is baked into the weights of both V3 and R1, even in the largest 671B parameter model. We're currently conducting analysis on the 671B model r
by pgkr 2y ago
Correct. The bias is baked into the weights of both V3 and R1, even in the largest 671B parameter model. We're currently conducting analysis on the 671B model running locally to cut through the speculation, and we're seeing interesting biases, including differences between V3 and R1.
Meanwhile, we've released the first part of our research including the dataset: https://news.ycombinator.com/item?id=42879698 https://news.ycombinator.com/item?id=42879698
- nicce 2y agoIs it really in the model? I haven’t found any censoring yet in the open models.
- lyu07282 2y agoIt isn't if you observe the official app it's API will sometimes even begin to answer before a separate system censors the output.
- homebrewer 2y agoReally? Local DeepSeek refuses to talk about certain topics (like Tiananmen) unless you prod it again and again, just like American models do about their sensitive stuff (which DeepSeek is totally okay with — I spent last night confirming just that). They're all badly censored which is obvious to anyone outside both countries.
- mmazing 2y agoNot my experience - https://imgur.com/xanNjun https://imgur.com/xanNjun just ran this moments ago.
- mmazing 2y agoWeird. Followup - I am getting censorship on the model from ollama's public model repository, but NOT from the models I got from huggingface running on a locally compiled llama.cpp.
- pgkr 2y agoYes, without a doubt. We spent the last week conducting research on the V3 and R1 open source models: https://news.ycombinator.com/item?id=42918935 https://news.ycombinator.com/item?id=42918935 Censoring and straight up propaganda is built into V3 and R1, even the open source version's weights.
- mmazing 2y agoI have not found any censorship running it on my local computer. https://imgur.com/xanNjun https://imgur.com/xanNjun
- pgkr 2y agoWe conducted further research on the full-sized 671B model, which you can read here: https://news.ycombinator.com/item?id=42918935 https://news.ycombinator.com/item?id=42918935 If you ran it on your computer, then it wasn't R1. It's a very common misconception. What you ran was actually either a Qwen or LLaMA model fine-tuned to behave more like R1. We have a more detailed explanation in our analysis.