Google DeepMind论文:4个Gemini
New Google DeepMind paper shows 4 general-purpose coding agents beating Minecraf…
多Agent自主协作解决长程复杂任务的实证研究,展示了无监督分工与共享记忆的潜力,对Agent架构设计有重要参考价值。
New Google DeepMind paper shows 4 general-purpose coding agents beating Minecraft with no human help, writing every program and splitting the work themselves.
Google DeepMind 最新论文显示,4 个通用编码代理在无人协助的情况下击败《我的世界》,自行编写每个程序并分工合作。
Finishing Minecraft means holding 1 plan across thousands of dependent steps, the same skill long software and business tasks need. Earlier Minecraft agents were trained on gameplay or handed game-specific planners.
通关《我的世界》意味着要在数千个相互依赖的步骤中保持一个统一的计划,这也是长期软件和企业任务所需的能力。此前的《我的世界》代理要么基于游戏玩法进行训练,要么被赋予特定于游戏的规划器。
Each Gemini 4 Argon agent got 1 goal, defeat the Ender Dragon, and a standard library that turns JavaScript into game actions. Nobody assigned roles or wrote game tools, so the team organized itself through chat and shared notes.
每个 Gemini 4 Argon 代理都获得了一个目标——击败末影龙,以及一个将 JavaScript 转换为游戏动作的标准库。没有人分配角色或编写游戏工具,因此团队通过聊天和共享笔记实现了自我组织。
From a random world on Easy difficulty, they killed the dragon in 27 hours and 14 minutes, writing 2,783 programs and dying 605 times. It was a single run, so how often this setup succeeds is still unknown.
在简单难度的随机世界中,它们在 27 小时 14 分钟内击杀了末影龙,共编写了 2,783 个程序,死亡 605 次。这是一次单次运行,因此该设置的成功率尚不明确。
Before writing custom planners for a long task, try general coding agents with a plain API and a shared memory.
在为长任务编写自定义规划器之前,可尝试使用普通 API 和共享内存的通用编码代理。
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