AI 与机器学习2024arxiv.org
评估生成模型所隐含的世界模型
Evaluating the world model implicit in a generative model
进阶 · 约 60 分钟读完前置知识：生成式模型基础、世界模型概念、探针/表征分析方法
导读
这篇论文考察生成式模型（如LLM）内部是否隐含地构建了关于世界的模型。研究者从模型的训练语料中提取地理知识，得到类似曼哈顿街道的地图表示，并用导航、最短路径和绕路等任务评估这些地图的准确度与鲁棒性——结果显示地图在常规路线上大致可用，但在需要规划绕路时会明显崩溃。评论提到训练用了8块A100。
HN 关注点：LLM是否真的形成了“世界模型”，讨论焦点是从模型提取的地图看似连贯却在绕路时失效，由此引发“世界模型”一词是否恰当、机器表征与人类认知差距的哲学争论。
HN 评论精选
这几条评论里：质疑 1 补充 2 亲历 1
质疑LLM确有内部表征，但称其为“世界模型”不恰当，因为它只基于词序列、无法建模真实世界的生成过程，与动物基于互动的预测模型有本质区别。
一个LLM必然要构建某种内部“模型”/表征来实现其“预测下一个词”的训练目标，因为要做好这件事需要相当深度和复杂的上下文识别能力，这可不是只看表面词序列的N-gram模型。然而，问题应该是：它构建了什么样的内部“模型”？把这种模型称为“世界模型”似乎很时髦，但在我看来并不恰当，而且它肯定与任何与真实世界互动、并从互动中学习的动物所构建的预测性表征大不相同。LLM是一个自回归模型——它试图仅凭词序列来预测训练样本的续写，并不了解那些词序列实际描述的世界，也无法建模创建那些样本的人类的生成过程，因为那个生成过程有感官输入等不同输入。人类或动物的“世界模型”是为了预测环境而构建的，但不是纯被动方式，而是由进化压力驱动、预测自身与环境互动的结果。
HarHarVeryFunny查看英文原文
An LLM necessarily has to create some sort of internal "model" / representations pursuant to its "predict next word" training goal, given the depth and sophistication of context recognition needed to to well. This isn't an N-gram model restricted to just looking at surface word sequences. However, the question should be what sort of internal "model" has it built? It seems fashionable to refer to this as a "world model", but IMO this isn't really appropriate, and certainly it's going to be quite different to the predictive representations that any animal that interacts with the world, and learns from those interactions, will have built. The thing is that an LLM is an auto-regressive model - it is trying to predict continuations of training set samples solely based on word sequences, and is not privy to the world that is actually being described by those word sequences. It can't model the generative process of the humans who created those training set samples because that generative process has different inputs - sensory ones (in addition to auto-regressive ones). The "world model" of a human, or any other animal, is built pursuant to predicting the environment, but not in a purely passive way (such as a multi-modal LLM predicting next frame in a video). The animal is primarily concerned with predicting the outcomes of it's interactions with the environment, driven by the evolutionary pressure to learn to act in way that maximizes survival and proliferation of its DNA. This is the nature of a real "world model" - it's modelling the world (as perceived thru sensory inputs) as a dynamical process reacting to the actions of the animal. This is very different to the passive "context patterns" learnt by an LLM that are merely predicting auto-regressive continuations (whether just words, or multi-modal video frames/etc).
补充对LLM能产出大致连贯的地图表示感到惊讶，同时指出论文未报告训练时长、数据集小易复现，以及Smalley模型更优却未解释。
尽管结果不对，我还是很佩服他们能从LLM中提取出任何看起来大致连贯的地图。最短路径图上有市中心和中央公园周边的一些街道并非全红，而中央公园本身在三张图上都很清晰。他们用了8块A100，但没说训练LLM花了多长时间，知道墙钟时间会很有意思。相对而言他们的数据集很小，意味着从零复现应该需要更少资源。不过有趣的是，Smalley模型表现更好，虽然他们没有推测原因。
fragmede查看英文原文
Wrong as it is, I'm impressed they were able to get any maps out of their LLM that look vaguely cohesive. The shortest path map has bits of streets downtown and around Central Park that aren't totally red, and Central Park itself is clear on all 3 maps. They used eight A100s, but don't say how long it took to train their LLM. It would be interesting to know the wall clock time they spent. Their dataset is, relatively speaking, tiny which means it should take fewer resources to replicate from scratch. What's interesting though is that the Smalley model performed better, though they don't speculate why that is.
补充从哲学角度类比：人类心智地图同样靠偶然互动构建，可能也存在虚构街道；NN地图常规导航近乎完美却无法绕路，令人反思人类地图的脆弱性。
我觉得这里有个哲学角度。我的世界地图是通过与真实世界的偶然互动构建的。这是否意味着我的世界地图对真实世界地图的接近程度，就像他们神经网络的地图对曼哈顿的接近程度一样？我的世界地图是否也充满不存在的街道、位置错误的出口？神经网络画的曼哈顿地图在用于正常导航时几乎100%正确，但在需要规划绕路时就会严重崩溃。我的世界地图有多脆弱？
isaacfrond查看英文原文
I think there is a philosophical angle to this. I mean, my world map was constructed by chance interactions with the real world. Does this mean that the my world map is a close to the real world map, as their NN's map is to Manhattan? Is my world map full of non-existent streets, exits that are at the wrong place, etc. The NN map of Manhattan works almost 100% correctly when used for normal navigation but breaks apart badly when it has to plan a detour. How brittle is my world map?
亲历分享将KGE侧车嵌入transformer的亲身实践：能做出稳定可交互的小模型，难点在于低成本更新知识而不重训，并看好transformer回归工具本位。
我见过一些非常令人印象深刻的结果：把预训练的知识图谱嵌入（KGE）模型嵌入transformer，让它“学习”去查询它（我只是在训练时用了异构损失函数，加上决定是否贪婪采样自KGE侧车的“分类器维度”，相信有更好的方法）。这显然只是主观看法，但我已经在这想法上玩了很多，很容易做出一个交互式的小LLM且结果稳定，唯一发现的问题是：如何低成本更新知识而不对LLM本身进行部分重训。不过对小型领域特定模型这不算问题——个人项目我用两块3090就行。我认为这些东西在transformer度过炒作曲线顶峰、成为构建特定类型模型的工具之后会变得有趣得多。
zxexz查看英文原文
I've seen some very impressive results just embedding a pre-trained KGE model into a transformer model, and letting it "learn" to query it (I've just used heterogenous loss functions during training with "classifier dimensions" that determine whether to greedily sample from the KGE sidecar, I'm sure there are much better ways of doing this.). This is just subjective viewpoint obviously, but I've played around quite a lot with this idea, and it's very easy to get a an "interactive" small LLM with stable results doing such a thing, the only problem I've found is _updating_ the knowledge cheaply without partially retraining the LLM itself. For small, domain-specific models this isn't really an issue though - for personal projects I just use a couple 3090s. I think this stuff will become a lot more fascinating after transformers have bottomed out on their hype curve and become a tool when building specific types of models.
相关资源
从这篇论文所有 HN 讨论帖的全部评论里整理出来的链接。
历次讨论
- 2024Hacker News 讨论帖159 分 · 45 评
同领域还值得读
DeepSeek-R1：通过强化学习激励 LLM 的推理能力
DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via RL
研究如何用强化学习激励大语言模型的推理能力，提出 DeepSeek-R1，并探讨 RL 训练对模型推理表现的影响。
AI 与机器学习2025 · 1,351 分 · 1,056 评 · 引用 30进阶 · 约 80 分钟编辑精选1-bit LLM 时代：面向低成本计算的三值参数
The Era of 1-bit LLMs: ternary parameters for cost-effective computing
提出用三值参数 {-1, 0, 1} 替代 LLM 中的浮点权重，即所谓“1-bit LLM”。
AI 与机器学习2024 · 1,040 分 · 447 评 · 引用 19进阶 · 约 45 分钟编辑精选DeepSeek-V3.2：推进开放大语言模型的前沿
DeepSeek-v3.2: Pushing the frontier of open large language models
这是 DeepSeek-V3.2 的技术报告，延续了 DeepSeek 开源权重模型的路线，目标是让开放模型继续逼近甚至追平前沿闭源模型。
AI 与机器学习2025 · 982 分 · 465 评进阶 · 约 90 分钟编辑精选通过自我对弈与通用强化学习掌握国际象棋和将棋（AlphaZero）
Mastering Chess and Shogi by Self-Play with General Reinforcement Learning
DeepMind 的 AlphaZero 论文：用通用强化学习算法（自我博弈 + MCTS + 神经网络），从零开始仅数小时训练便在国际象棋和将棋上击败了世界顶级引擎 Stock…
AI 与机器学习2017 · 539 分 · 270 评 · 引用 13进阶 · 约 90 分钟编辑精选用强化学习发现更快的矩阵乘法算法
Discovering faster matrix multiplication algorithms with reinforcement learning
论文提出 AlphaTensor，用强化学习搜索矩阵乘法算法。
AI 与机器学习2022 · 518 分 · 111 评专业 · 约 90 分钟编辑精选