HackerNews中文版

朋友们，我有一个想法，可以使神经网络以确定性的方式而不是概率性的方式运行。目前，我们在非结构化数据上训练神经网络，但问题是它们是概率模型，难以理解。我想创建一个全新的神经网络，它完全可以理解，并且每个权重都是有意的。所以我们要做的是这样的。我们创建一个专门用于读取、理解和编写实际神经网络权重的神经网络。所以这个想法是，一个训练有素的网络实际上可以有意地创建和更新神经网络，并以确定性的方式获得新知识。例如，你可以说“创建一个可以读取 MNIST 的网络”，它实际上会知道如何创建网络，包括权重，并且它会为网络中的权重提出合理的值。它将指定每个神经元和每个连接，并给出权重的实际值。这很酷的一点是，我们可以让它对各种神经架构获得直觉。它会设置权重，分配输入值，运行它，然后“调试”，这样它在创建神经网络方面就会变得越来越好。说实话，我们可以让它像进行强化学习一样，每当它做出更好的更新时，它就会说“是”！并通过对该系列进行强化学习来奖励自己，我们并行进行此操作，并且成功的获胜。因此，好处在于对安全性敏感的场景，拥有一个能够真正理解和检查权重意味着什么以及它们用途的 AI，以及能够为了特定目的编辑它们的能力。例如，这将防止手术机器人因为在预训练中看到了关于屠宰场的视频而留下训练不佳的神经元，导致切除过多。另一个好处是它可以像我们的大脑一样智能地工作——从早期层到后期层创建“跳跃”连接，从而提高效率。它还可以提高效率，因为它只创建必要的几个连接。它还可以智能地选择数据类型，对敏感且需要高精度浮点数的区域使用高精度浮点数，而对其他区域使用低精度浮点数。通过训练一个网络来检查和创建网络，我们可以更接近于保证网络没有流氓神经元。由于网络有数十亿个神经元，我猜它需要逐位地对神经元进行高级别和低级别的检查，并对不同的部分进行大量的工作和实验，并创建不同部分指代内容的纯文本“数据库”，它可以创建索引等。最终，一个神经网络可以像一种语言一样“自我编译”，甚至不需要预训练阶段或反向传播。

查看原文

Hey friends I have an idea for a way to make neural networks deterministiclly instead of probabilistically.Right now we train neural networks on unstructured data but the problem is they are probabilistic models and hard to understand.I want to create a new neural network that is fully understandable, and each weight is intentional.So what we do is this. We create a neural network that specializes in reading, understanding and writing actual neural network weights.So the idea is that a correctly trained network could actually intentionally create and update neural networks with new knowledge deterministically.So like you could say "create a network that can read mnist", and it would like actually know how to make the network including the weights, and it would come up with reasonable values for the weights in the network. It would specify each nerve and each connection with the actual value of the weight.The cool thing about this is we could have it like gain an intuition for various neural architectures. it would set weights, assign input values, run it through, and "debug" so it would get better and better at making neural nets.Honestly, we could have it like do reinforcement learning where every time it makes updates that are better it can be like "yes"! and it will reward itself by doing reinforcement learning for that series, and we do this in parallel and the ones that work win.So the benefit would be for safety sensitive scenarios, having an ai that can truly understand and inspect what weights mean and what they are for, and ability to edit them for precise purpose. This would prevent a surgery robot, for example, from having neurons left over poorly set from bad training cutting too much because it saw a video in pretraining about a butcher shop.The other benefit is that it could intelligently do what our brains do - create "skip" connections from early layers to later layers enhancing efficiency.It could also lead to enhanced efficiency where it only makes a few connections that are necessary. It could also choose data types intelligently using high precision floating points for areas that are sensitive and need it, and low precision elsewhere.By training a network to be able to inspect and make networks, we can get much closer to guaranteeing that networks don't have rogue neorons.Since networks have billions of neurons, I would guess that it would need to do inspection of neurons at high levels and low levels bit by bit and a ton of work and experimentation on different sections and sort of create a plain text "database" of what sections refer to what, it could make indexes and stuff like that.Eventually a neural network could be "self compiling" like a language where it doesn't even need a pretraining phase or backprop.

安全人工智能