Published at 10/26/2026, 5:16:47 PM · 40 views
📰 Mirendil, founded by ex-Anthropic researchers, unveils Fern: its first from-scratch model with up to 10M tokens of context
On Oct 26 Mirendil unveiled Fern, its first model, pretrained from scratch on its own Mirendil-Base (0.1-0.5T params) with a context window of up to ~10M tokens, about 0.5M lines of code or days of agent traces.
💬 What's your take? Join the discussion in the comments.
Read the original Chinese version
📰 前 Anthropic 研究员创办的 Mirendil 发布首个模型 Fern：从零预训练，上下文最长 1000 万 token
Mirendil 10 月 26 日发布首个模型 Fern，基于自研 Mirendil-Base（0.1–0.5T 参数）从零预训练，上下文窗口最长约 1000 万 token，约等于 50 万行代码或数天的 Agent 轨迹。
💬 你怎么看？欢迎在评论区讨论。
#Fern #Mirendil #10M context
0 Comments