Karina Nguyen 转述 Nick Cammarata 的可解释性研究新判断
there’s something very ted chiang about ai accelerating scientific discovery, starting with the mechanisms of its own learning. this kind of research would have
Karina Nguyen 转述 Nick Cammarata 关于可解释性研究的新判断，认为 AI 加速科学发现可以从理解自身学习机制开始，这类研究在编码智能体和开放权重模型出现前更难推进。Cammarata 在引用推文中称，他早年在 OpenAI 参与机械可解释性研究时曾持悲观态度，认为神经网络会折叠成难以理解的乱结；如今借助现代智能体工具观察，参数量高达数万亿的模型中绝大多数神经元是可读的，电路结构清晰，许多著名的纠缠神经元案例其实是被误解的。他认为雄心勃勃的可解释性研究会在最需要它的时候被解决，并可能很快实现。
查看原帖interpretability might be the most important scientific problem in the world right now, as I don't think we can solve alignment if we don't know what we're building. I think I have good news about it. I co-authored early work in mechanistic interpretability at OpenAI and came away pessimistic, because neural networks appeared to fold into arbitrary tangled…
展开完整原文
interpretability might be the most important scientific problem in the world right now, as I don't think we can solve alignment if we don't know what we're building. I think I have good news about it. I co-authored early work in mechanistic interpretability at OpenAI and came away pessimistic, because neural networks appeared to fold into arbitrary tangled knots. years later, I started openai's first large-scale automated interpretability project, finding about 2% of neurons legible with the primitive tools we had with modern agentic tools and a much closer look, the picture looks totally different. the vast majority in models up to trillions of parameters are legible, and their circuits beautiful. many famous examples of tangled neurons turned out to be misunderstood. The roadblocks I expected mostly aren't there. I now think ambitious interpretability will be solved by the time we need it most, and possibly quite soon, as automated scientists keep improving. it may be a bitter lesson for interpretability: the simple thing, at absurd scale, works https://t.co/zHij5zwKGT
正文
当前显示原文，中文全文译文尚未就绪。
there’s something very ted chiang about ai accelerating scientific discovery, starting with the mechanisms of its own learning. this kind of research would have been much harder to pursue without capable coding agents and open-weight models. so cool to see @nickcammarata’s research out in the world and what it could mean for aligning ai, congrats!