Agent S4 通过自主进化界面设计刷新开源计算机操作智能体纪录
Evolving Agent Interfaces
Simular AI 的 Agent S4 自己学着给智能体设计操作界面，用 Qwen3.8-27B 在 OSWorld V2 上跑赢了 GPT-5.5 和 Claude Opus 4.8，成本还只有零头。
Simular AI 推出 Agent S4，让一个 meta-agent 分析任务智能体的执行轨迹，自动构建并修订 Evolving Agent Interfaces 仓库，再通过基于种群的选择保留能泛化到未见任务的设计。在 OSWorld V2 上，Agent S4 搭配 Qwen3.8-27B 取得 21.3% binary score，是固定界面下同模型的 3.8 倍，超过 GPT-5.5 的 13.0% 和 Claude Opus 4.8 的 20.6%，成本分别只有约 4.3% 和 1.5%。在 ScienceBoard 基准上，进化后的界面把同模型成功率从 41.4% 提升到 58.6%。代码已在 GitHub 开源。
Evolving Agent Interfaces
Computer-use agents are limited by the interfaces humans give them. An agent may type a spreadsheet cell by cell when a single line of code would do, or print all two thousand lines of a file through a shell to find one function. Both the interfaces and the knowledge of when to use each are built by hand, which cannot keep pace with new applications, workflows, and models. We argue that interface design should itself be an agent capability. We introduce Agent S4, an autonomous computer-use scientist that builds Evolving Agent Interfaces from task experience. A meta-agent examines task agents' trajectories on synthetic training tasks, without success labels, and constructs and revises interfaces and usage knowledge in a repository. Population-based evolution then selects which repositories to develop further by their performance on validation tasks the meta-agent never sees, favoring changes that generalize beyond the seen tasks. On OSWorld V2, Agent S4, with Qwen3.8-27B and the evolved repository of interfaces, sets a new state of the art in open-source computer-use agents, reaching 21.3% binary score, 3.8 times the same model with fixed interfaces. This is above both GPT-5.5 (13.0%) and Claude Opus 4.8 (20.6%), at about 4.3% and 1.5% of their respective costs. On ScienceBoard, the evolved interfaces improve the success rate from 41.4% to 58.6% with the same model. These results suggest that useful interfaces can be evolved from experience, making autonomous interface design a source of improvement complementary to advances in models and harnesses. We open-source our code at https://github.com/simular-ai/agent-s.