Anthropic 可解释性研究员 Emmanuel Ameisen 谈 Claude 内部机制研究
"To make them is not to understand them." Emmanuel Ameisen (@mlpowered) has spent the last 3 years ...
Anthropic 研究员聊他怎么打开 Claude 看 AI 到底在想什么，还发现模型会提前规划好几个 token，挺有意思
Emmanuel Ameisen 在 Anthropic 做了 3 年可解释性研究，专门分析 Claude 等 LLM 的内部运作。他提出模型是"长"出来的而不是搭建出来的，并发现模型会提前规划多个 token 的输出。访谈还涉及拟人化 AI 为何可能是合理方法、"J-space"如何揭示模型思考方式，以及可解释性与 AI 对齐安全之间的关联。
"To make them is not to understand them." Emmanuel Ameisen (@mlpowered) has spent the last 3 years ...
"To make them is not to understand them." Emmanuel Ameisen ( @mlpowered ) has spent the last 3 years doing interpretability research at @AnthropicAI , looking inside LLMs like Claude to figure out what they're actually doing. As he puts it, models are "grown," not built. We discuss: 🔸 His discovery that models plan many tokens ahead 🔸 Why anthropomorphizing AI can actually be the right approach 🔸 What the "J-space" reveals about how models think 🔸 How interpretability connects to AI alignment and safety Listen no youtu.be/JHo-aiePjoM fpkab 💬 14 🔄 0 ❤️ 9 👀 3866 📊 14 ⚡
