Reflection发布Beam模型:501B参数,强调推理效率
Reflection has announced Beam: a 501B-parameter model with 23B active, with full…
Reflection has announced Beam: a 501B-parameter model with 23B active, with full weights scheduled for release this month. Frontier capabilities, and holy is this model efficient!
Reflection 宣布推出 Beam:一个拥有 501B 参数、23B 激活参数的模型,完整权重计划于本月发布。具备前沿能力,而且该模型的效率令人惊叹!
Trained from scratch, Beam targets coding, reasoning and agentic tasks.
Beam 从零开始训练,专注于编码、推理和智能体任务。
Reflection claims 3–4x higher inference efficiency than GLM 5.2 and over 4x higher efficiency than leading Western open models in its class.
Reflection 声称其推理效率比 GLM 5.2 高出 3–4 倍,在同类领先西方开源模型中效率高出 4 倍以上。
24T pretraining tokens in four weeks, followed by four weeks of reinforcement learning on 10,500 GB300s.
四周内完成 24T 预训练 token,随后在 10,500 台 GB300 上进行四周的强化学习。
Terminal Bench v2.1 and 80.9 on SWE-bench Verified.
Terminal Bench v2.1 得分 80.9,SWE-bench Verified 得分 80.9。
But again: look at the efficency, eypecially in Terminal bench v2.1! Congrats on that launch. More western open models to come.
但再次强调:请关注效率,尤其是在 Terminal Bench v2.1 上!祝贺此次发布。更多西方开源模型即将推出。
更进一步:量化金融体系
看懂新闻只是起点——沿量化金融路径,把它变成能交付的工程能力