迷你激活实现最大真实智能，模型仅用98亿参数每词激活却达顶尖性能。
The MiniMax-M2 Series: Mini Activations Unleashing Max Real-World Intelligence
- 采用极小激活机制，每词仅激活98亿参数，大幅降低计算开销。
- 在编程、搜索、办公任务等基准上达到前沿水平，效率与性能兼备。
- 支持自主调试与自我演进，适合需要高效智能代理的场景。
我们推出MiniMax-M2系列，一类基于‘极小激活释放最大真实智能’理念的专家混合语言模型。旗舰模型M2总参数量为2299亿，每词仅激活98亿参数。该系列从零开始设计用于智能体部署，包含三大组件：(i) 由智能体驱动的数据流水线，生成大规模可验证轨迹，涵盖智能体编程与协作任务，基于可执行工作区与对齐奖励；(ii) Forge，一个可扩展的原生智能体强化学习系统，适配长周期智能体轨迹，结合窗口化FIFO调度、前缀树合并、推理优化及训练-推理-智能体解耦设计，支持白盒与黑盒智能体；(iii) 最新M2.7检查点已初步实现自演化——能自主调试训练过程并修改自身架构。从M2到M2.7，该组合将极小激活优势转化为在智能体编程、深度搜索、办公任务和推理基准上的顶级表现。
原文摘要 · Abstract (English)
We introduce the MiniMax-M2 series, a family of Mixture-of-Experts language models built around the principle that mini activations can unleash maximum real-world intelligence. The flagship M2 contains 229.9B total parameters with only 9.8B activated per token. Designed end-to-end for agentic deployment, the M2 series rests on three components: (i) agent-driven data pipelines producing large-scale, verifiable trajectories across agentic coding and agentic cowork, each grounded in an executable workspace and an artifact-aligned reward; (ii) Forge, a scalable agent-native RL system that adapts to long-horizon agent trajectories, paired with windowed-FIFO scheduling, prefix-tree merging, inference optimization, and a clean training-inference-agent decoupling that supports both white-box and black-box agents; (iii) the latest M2.7 checkpoint takes an early step toward self-evolution -- autonomously debugging training runs and modifying its own scaffold. Across M2 through M2.7, this combination translates a mini-activation footprint into frontier-tier performance on agentic coding, deep search, office-task, and reasoning benchmarks.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。