用能量视角统一机器学习：一篇 arXiv 教程把 GAN、diffusion、BERT 都看作 EBM
An Energy-Based Perspective on Machine Learning
做 ML 研究或想搞懂各模型之间关系的人值得花时间。它把 GAN、diffusion、BERT、对比学习全塞进同一个能量框架，差别只在配分函数和采样怎么处理。看懂这篇，很多论文里的公式其实是在说同一件事，而且能量组合还能直接推理出新算法思路。
Yann LeCun 等人在 arXiv 发布教程《An Energy-Based Perspective on Machine Learning》（arXiv:2611.14383），提出用能量模型（EBM）作为统一框架解释整个机器学习领域。EBM 通过标量能量函数定义分布，即 p_θ(x)∝exp(-E_θ(x))，任何一个 EBM 都由三个选择决定：能量函数、推断过程和学习方法。
教程的核心论证是：判别模型、GAN、normalizing flows、diffusion 与自回归模型、对比学习、word2vec、BERT、去噪自编码器、带 EBM 策略的强化学习、逆强化学习、规划、以及带验证器的推理时计算，本质上都是 EBM，区别只在于如何处理配分函数和采样。文中梳理了最大似然、噪声对比估计（NCE）、伪似然、score matching、对抗训练等方法，指出它们共享同一个对比结构。
这个视角的实际价值在于：把现有方法纳入统一框架后，可以通过替换三个组件中的一个来构造新算法，而能量的组合（composition of energies）提供了在推理时组合多个学习模型的原理性方法。对研究者而言，这是一份把看似分散的模型谱系打通的系统性材料。
An Energy-Based Perspective on Machine Learning
Energy-based models (EBMs) define a distribution through a scalar energy, $p_θ(x)\propto\exp(-E_θ(x))$, and are usually presented as one model family among many. This tutorial has a broader goal: beyond introducing EBMs, it uses them as a single lens that connects the different aspects of machine learning. Any EBM is specified by three choices, the energy, which determines what the model represents; the inference procedure, which samples from or optimizes the energy; and the learning method, which fits the energy to data. We first develop these components, including sampling, variational inference, and learning by maximum likelihood, noise contrastive estimation, pseudolikelihood, score matching, and adversarial training, and show that the learning methods share one contrastive structure. We then show that discriminative models, generative models such as GANs, normalizing flows, diffusion and autoregressive models, representation learning methods such as contrastive learning, word2vec, BERT, denoising auto-encoders, and diffusion representations, reinforcement learning with EBM policies and world models, inverse reinforcement learning, and planning, as well as inference-time computation with verifiers, are all EBMs that differ in these three choices, mostly in how they handle the partition function and sampling. This view explains existing methods, organizes them into a common framework, and suggests new algorithms by replacing one component while keeping the others, with composition of energies as a principled way to combine learned models at inference time.
本条情报归属 AITOP「论文」核心栏目，重点聚焦北美前沿实验室与硅谷科技巨头。在跨区域产业观察坐标中，该动态直接关联「AI 开发者工具链、Coding Agent 与工程化实践」研究主线，反映出当前技术演进与商业落地的最新风向。涉及核心实体与技术议题包括：能量模型、diffusion、BERT、GAN，读者可通过下方标签进入对应专题页查阅历史脉络与关联对比。
信源核验说明：本文原始线索采编自「arXiv cs.AI」（@Bo Dai, Yilun Du），经 AITOP 多维评分管线结构化提炼。
