Andrej Karpathy 相信法学硕士将使 PyTorch 等深度学习框架变得多余

法学硕士正在成为企业和研究人员面临的许多问题的默认解决方案。即使涉及语言和文本之外的领域，人们也一直在尝试使用 LLM 来猜测和预测下一个标记，这引发了围绕 PyTorch 等其他工具的需求的有趣对话，因为 LLM 将来可以完成全部工作。

The rapid advancement of large language models (LLMs) has sparked a discussion about their potential to replace other tools, such as deep learning frameworks like PyTorch, for a wide range of problems.

大型语言模型 (LLM) 的快速发展引发了关于它们是否有可能取代其他工具（例如 PyTorch 等深度学习框架）来解决各种问题的讨论。

LLMs are typically designed to predict the next token in a sequence, whether that sequence consists of words, images, or other types of information. This next token prediction framework can be applied to a diverse set of problems, extending beyond text.

LLM 通常旨在预测序列中的下一个标记，无论该序列由单词、图像还是其他类型的信息组成。下一个令牌预测框架可以应用于多种问题，扩展到文本之外。

Andrej Karpathy suggests that deep learning frameworks, like PyTorch and its counterparts, might be overly general for the majority of problems in the future.

Andrej Karpathy 认为，像 PyTorch 及其对应产品这样的深度学习框架对于未来的大多数问题可能过于通用。

LLMs have evolved beyond their initial specialization in language. The term “language” is used historically because these models were first trained to predict the next word in a sentence. However, LLMs can work on any kind of data that’s broken down into small pieces, called tokens.

法学硕士已经超越了最初的语言专业。 “语言”一词在历史上一直被使用，因为这些模型首先被训练来预测句子中的下一个单词。然而，法学硕士可以处理任何类型的数据，这些数据被分解成小块，称为令牌。

Imagine LLMs as a super-smart guessing game. If you’re building a car, a house, or an animal using Legos, you’re just putting blocks together. LLMs don’t care if the tokens (blocks) represent words, images, or even molecules—they just focus on predicting what the next block should be based on what’s already there.

将法学硕士想象成一个超级智能的猜谜游戏。如果您使用乐高积木建造汽车、房屋或动物，您只需将积木拼凑在一起即可。法学硕士不关心标记（块）是否代表单词、图像甚至分子，他们只是专注于根据已有的内容预测下一个块应该是什么。

For example, protein prediction models like AlphaFold and ESMFold are built on top of generative language models. Calling such intricate models LLMs might be limiting.

例如，AlphaFold 和 ESMFold 等蛋白质预测模型是建立在生成语言模型之上的。将如此复杂的模型称为法学硕士可能是有局限性的。

Karpathy further highlights that the term “language” might be leading people to believe that LLMs are limited to text applications, which is not the case.

Karpathy 进一步强调，“语言”一词可能会让人们认为法学硕士仅限于文本应用程序，但事实并非如此。

“I don’t think this is true but I think it’s half true,” Karpathy said in response to a thread about LLMs being able to handle almost all types of problems.

“我不认为这是真的，但我认为这只是对了一半，”卡帕蒂在回应有关法学硕士能够处理几乎所有类型问题的帖子时说道。

“Probably the name should change.”

“也许这个名字应该改变一下。”

“Definitely needs a new name. ‘Multimodal LLM’ is extra silly, as the first word contradicts the third word,” replied Elon Musk in the same thread.

“肯定需要一个新名字。 “多模式法学硕士”特别愚蠢，因为第一个词与第三个词相矛盾，”埃隆·马斯克在同一条帖子中回答道。

Meanwhile, Yann LeCun is more concerned about why this doesn’t make sense for all the types of problems.

与此同时，Yann LeCun 更关心的是为什么这对所有类型的问题都没有意义。

“It only works with discretized outputs (discrete symbols) and only makes sense with symbol sequences with a natural order (not images). Text, DNA, proteins, musical scores, etc. are discrete or easily discretized,” said LeCun.

“它仅适用于离散输出（离散符号），并且仅对具有自然顺序的符号序列（而非图像）有意义。文本、DNA、蛋白质、乐谱等都是离散的或容易离散的，”LeCun 说。

For something like images, which are continuous and don’t naturally have a strict sequence of discrete symbols (each pixel doesn’t follow a clear ‘order’ like text), LLMs don’t work as naturally. To use an LLM for images, you would first need to somehow convert the image into discrete chunks (like dividing the image into small patches), but this doesn’t follow the same natural order that exists in text or DNA.

对于像图像这样的东西，它们是连续的，并且自然不具有严格的离散符号序列（每个像素不遵循像文本那样的明确“顺序”），法学硕士不能自然地工作。要将 LLM 用于图像，您首先需要以某种方式将图像转换为离散的块（例如将图像划分为小块），但这并不遵循文本或 DNA 中存在的相同自然顺序。

Agreeing with Karpathy, and a little with LeCun, Gary Marcus said that statistical modelling of token streams works well if reasoning or planning isn’t required.

Gary Marcus 同意 Karpathy 的观点，也同意 LeCun 的观点，他表示，如果不需要推理或规划，令牌流的统计建模效果很好。

Last month, Eliezer Yudkowsky also said that predicting the next token can solve almost all the well-posed problems.

上个月，Eliezer Yudkowsky 还表示，预测下一个代币可以解决几乎所有适定问题。

“literally any well-posed problem is isomorphic to ‘predict the next token of the answer’,” he said.

“实际上，任何适定问题都同构于‘预测答案的下一个标记’，”他说。

“Throwing an LLM at it”

“攻读法学硕士”

The idea that many problems can be reduced to a token-stream prediction model is intriguing, especially since domains like images, audio, and even molecules can be broken down into sequences of tokens. This suggests that a unified approach like LLMs could handle diverse tasks, reducing the need for highly specialised architectures, such as PyTorch.

许多问题可以简化为令牌流预测模型的想法很有趣，特别是因为图像、音频甚至分子等领域可以分解为令牌序列。这表明像法学硕士这样的统一方法可以处理不同的任务，从而减少对 PyTorch 等高度专业化架构的需求。

However, frameworks like PyTorch provide more than just flexibility in creating neural network models. They allow for a variety of deep learning operations that aren’t necessarily relevant for LLMs but are critical for other areas like reinforcement learning, generative models, and non-sequential tasks.

然而，像 PyTorch 这样的框架不仅仅提供创建神经网络模型的灵活性。它们允许进行各种深度学习操作，这些操作不一定与法学硕士相关，但对于强化学习、生成模型和非序列任务等其他领域至关重要。

While it’s true that LLMs could dominate many applications, not every problem is best framed as “next token prediction.” We may see a simplification or specialization of deep learning frameworks to accommodate the increasing dominance of LLM-based models. Still, the complete redundancy of frameworks like PyTorch might be too extreme of a prediction.

虽然法学硕士确实可以主导许多应用程序，但并不是每个问题都最好被定义为“下一个代币预测”。我们可能会看到深度学习框架的简化或专业化，以适应基于法学硕士的模型日益占主导地位。尽管如此，像 PyTorch 这样的框架的完全冗余可能是一个过于极端的预测。

OpenAI’s newest model o1 gives a sense of why LLMs would be able to solve a lot of problems outside of the realm that is currently considered achievable. With the reasoning tokens in place, the model can go beyond just ‘predicting’ the next token and giving reasons for why it did so.

OpenAI 的最新模型 o1 让人了解为什么法学硕士能够解决目前认为可以实现的领域之外的许多问题。有了推理标记，模型就不仅仅是“预测”下一个标记并给出这样做的理由。

PyTorch and similar frameworks may not become redundant but could evolve to become more focused on token-based models, while still offering tools for more diverse problems outside that paradigm.

PyTorch 和类似的框架可能不会变得多余，但可能会发展为更加关注基于代币的模型，同时仍然为该范式之外的更多样化的问题提供工具。

Though, currently only in language or text format, the capabilities might extend beyond it soon. Calling LLMs as LLMs might be underrepresenting their capabilities. Moreover, “it just predicts the next token” is a thought-terminating cliche.

虽然目前仅采用语言或文本格式，但功能可能很快就会超越它。将法学硕士称为法学硕士可能低估了他们的能力。此外，“它只是预测下一个标记”是一个终结思想的陈词滥调。