Inherent 称小模型 AI 智能体论文复现表现超 Anthropic 与 OpenAI
Inherent 用极小模型证明，强化学习能让 AI 在科研复现任务上超越巨头，为 AI 科学家之路探路。
Inherent 是一家由 DeepMind 校友在伦敦创立的 AI 实验室，其新发布的 AI 智能体 Faraday 在独立复现科研论文结论的任务上，超越了 Anthropic 和 OpenAI 的更大模型。Faraday 基于仅 270 亿参数的模型，并通过强化学习（一种奖励好结果而非逐步指令的训练方法）来培养“研究品味”。
正文摘录
[Inherent](https://inherentlabs.ai/), a London AI lab founded by [Google DeepMind alumni](https://tech.eu/2026/05/01/over/), says its AI agent just outperformed much larger models from Anthropic and OpenAI using a fraction of the size. Of all the startups launched by Google DeepMind alumni, Inherent has gotten relatively little attention. But while better-funded rivals have yet to show the world anything concrete, the London-based team is starting to share what it’s been building. Just weeks after emerging from stealth with a [$50 million seed round](https://tech.eu/2026/05/29/london-based-ai-lab-inherent-emerges-from-stealth-with-50m-raise/), the British startup says its newly released AI agent, [Faraday](https://inherentlabs.ai/research/training-to-replicate), has [outperformed](https://…