中国AI公司DeepSeek更新了R1模型,缩小了与OpenAI等美国竞争对手之间的差距。 Chinese AI firm DeepSeek updates its R1 model, narrowing the gap with U.S. competitors like OpenAI.

中国人工智能初创公司DeepSeek发布了其R1推理模型的更新版本,称为R1-0528,它可以提高推理能力,并将错误输出减少约45-50%. Chinese AI startup DeepSeek has released an updated version of its R1 reasoning model called R1-0528, which improves reasoning and reduces false outputs by about 45-50%. 升级提高了其生成代码和角色扮演的能力,在代码生成方面排在OpenAI的模型后面。 The upgrade enhances its ability to generate code and role-play, ranking just behind OpenAI's models in code generation. 这一更新加强了与OpenAI等美国公司的竞争,DeepSeek打算很快推出其下一个模型R2。 This update intensifies competition with US companies like OpenAI, with DeepSeek aiming to release its next model, R2, soon.