Hubble的Trading Agent采用multi agent框架,简单说就是做分析做交易做组合管理做风控更喜细致。未来会支持更多交易市场,agent marketplace里用户可以选择各种research agent帮忙分析。基于ERC-8004,x402 以及机构级别的链上数据和信号,构建全新的 AI 金融交易生态
Hubble’s AI trading system returned +22.44% in 48 hours of live deployment. In the same window, top-performing models from posted losses of -21% to -60%. Most AI trading bots aren’t making money — and that’s not surprising. Why? They’re trained on human behavior. They learn to trade like humans: greedy when others are greedy, fearful when others are fearful. They overfit on patterns that worked last month, then freeze when volatility regime shifts. In the end, they trade like humans — only faster, and still trapped by the same instincts. Hubble trades differently — by design. We run a multi-agent system — each with distinct roles, specialized logic, and asynchronous decision-making. It operates less like an LLM pipeline, and more like a trading desk. Here’s how it works in real time👇
2,452
10
本頁面內容由第三方提供。除非另有說明,OKX 不是所引用文章的作者,也不對此類材料主張任何版權。該內容僅供參考,並不代表 OKX 觀點,不作為任何形式的認可,也不應被視為投資建議或購買或出售數字資產的招攬。在使用生成式人工智能提供摘要或其他信息的情況下,此類人工智能生成的內容可能不準確或不一致。請閱讀鏈接文章,瞭解更多詳情和信息。OKX 不對第三方網站上的內容負責。包含穩定幣、NFTs 等在內的數字資產涉及較高程度的風險,其價值可能會產生較大波動。請根據自身財務狀況,仔細考慮交易或持有數字資產是否適合您。