Claude3-Opus
Claude3-Opus is an AI model published by Anthropic, released on 2024-03-04, for 多模态大模型, with 0.0B parameters, and 200K tokens context length, under the 不开源 license.
Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology
Claude3-Opus currently shows benchmark results led by GSM8K (8 / 26, score 95), HellaSwag (1 / 3, score 95.40), MMLU (26 / 64, score 86.80). This page also consolidates core specs, context limits, and API pricing so you can evaluate the model from benchmark results and deployment constraints together.
Claude3-Opus是Anthropic公司发布的第三代多模态大语言模型。第三代的Claude-3模型包含3个版本,这里说的Claude3-Opus是其中能力最强的模型。各项评测人任务结果都非常好,甚至超过了GPT-4。
在多模态方面,Claude3-Opus也有强大的能力。

Claude2最受诟病的就是无效的拒绝回答。由于Anthropic在对齐方面做了严格的工作,导致Claude2.1经常出现拒绝回答的情况。在Claude3-Opus上。Anthropic做了改进,在内部测试中,Claude2.1错误地拒绝比例大概在26%左右,而Claude3-Opus上这个比例下降到了11%,进步明显!

欢迎关注 DataLearner 官方微信,获得最新 AI 技术推送
