Characterizing the implicit structure of the computation within neural networks is a foundational problem in the area of deep learning interpretability. Can their inner decision process be captured symbolically in some familiar logic? We show that any transformer neural network can be translated into an equivalent fixed-size first-order logic formula which may also use majority quantifiers. The idea is to simulate transformers with highly uniform threshold circuits and leverage known theoretical connections between circuits and logic. Our findings also reveal the surprising fact that the entire transformer computation can be reduced merely to the division of two (large) integers. While our results are most pertinent for transformers, they apply equally to a broader class of neural network architectures, namely those with a fixed-depth uniform computation graph made up of standard neural net components, which includes feedforward and convolutional networks.

语言模型通常仅在文本上进行培训，而无需其他基础。关于从这种过程中可以推断出多少自然语言语义的争论。我们证明，可以从理想的语言模型中提取句子之间的判断，该模型可以完美地了解其目标分布，假设训练句子是由Gricean Agents产生的，即遵循实用学语言学理论的基本交流原理的代理人。我们还表明，可以从对这种Gricean数据训练的语言模型的预测中解码需要判断。我们的结果揭示了一种理解未标记的语言数据中编码的语义信息的途径，以及从语言模型中提取语义的潜在框架。

我们证明，可以通过恒定的深度统一阈值电路模拟输入长度中具有对数精度的变压器神经网络（以及使用输入长度中的线性空间计算的FeedForward子网络）。因此，此类变压器仅在$ \ mathsf {tc}^0 $中识别形式语言，这是由常数深度，多大小阈值电路定义的语言类。这证明了NLP中的实际主张与计算复杂性理论中的理论猜想之间的联系：“注意就是您需要的一切”（Vaswani等，2017），即，只有在所有有效地计算的情况下，变形金刚都能够进行所有有效的计算可以使用日志空间来解决问题，即$ \ mathsf l = \ mathsf p $。我们还构建了一个可以在任何输入上评估任何恒定深度阈值电路的变压器，证明变形金刚可以遵循$ \ Mathsf {tc}^0 $中表示的说明。

最近在NLP中的工作已经记录了输入功能和输出标签之间的DataSet工件，偏置和虚假相关性。但是，如何判断哪些功能具有“虚假”而不是合法相关性通常留下未指定。在这项工作中，我们认为，对于复杂的语言理解任务，所有简单的特征相关性都是虚假的，我们将这一概念正式化为一类我们称之为能力问题的问题。例如，自己的“惊人”一词不应提供关于情绪标签的信息，无论出现的背景，哪些内容都可以包括否定，隐喻，讽刺等。我们理论上分析创建能力问题数据的难度当考虑人类偏见时，显示现实数据集将越来越偏离能力问题，因为数据集大小增加。此分析为我们提供了一个简单的数据集工件统计测试，我们用于显示比在事先工作中描述的更细微的偏见，包括展示模型与这些不太极端的偏差影响不恰当地影响。我们对此问题的理论处理也允许我们分析所提出的解决方案，例如将本地编辑为数据集实例制作，并为未来的数据收集和模型设计努力提供目标能力问题的建议。

We present Muse, a text-to-image Transformer model that achieves state-of-the-art image generation performance while being significantly more efficient than diffusion or autoregressive models. Muse is trained on a masked modeling task in discrete token space: given the text embedding extracted from a pre-trained large language model (LLM), Muse is trained to predict randomly masked image tokens. Compared to pixel-space diffusion models, such as Imagen and DALL-E 2, Muse is significantly more efficient due to the use of discrete tokens and requiring fewer sampling iterations; compared to autoregressive models, such as Parti, Muse is more efficient due to the use of parallel decoding. The use of a pre-trained LLM enables fine-grained language understanding, translating to high-fidelity image generation and the understanding of visual concepts such as objects, their spatial relationships, pose, cardinality etc. Our 900M parameter model achieves a new SOTA on CC3M, with an FID score of 6.06. The Muse 3B parameter model achieves an FID of 7.88 on zero-shot COCO evaluation, along with a CLIP score of 0.32. Muse also directly enables a number of image editing applications without the need to fine-tune or invert the model: inpainting, outpainting, and mask-free editing. More results are available at https://muse-model.github.io

We introduce Argoverse 2 (AV2) - a collection of three datasets for perception and forecasting research in the self-driving domain. The annotated Sensor Dataset contains 1,000 sequences of multimodal data, encompassing high-resolution imagery from seven ring cameras, and two stereo cameras in addition to lidar point clouds, and 6-DOF map-aligned pose. Sequences contain 3D cuboid annotations for 26 object categories, all of which are sufficiently-sampled to support training and evaluation of 3D perception models. The Lidar Dataset contains 20,000 sequences of unlabeled lidar point clouds and map-aligned pose. This dataset is the largest ever collection of lidar sensor data and supports self-supervised learning and the emerging task of point cloud forecasting. Finally, the Motion Forecasting Dataset contains 250,000 scenarios mined for interesting and challenging interactions between the autonomous vehicle and other actors in each local scene. Models are tasked with the prediction of future motion for "scored actors" in each scenario and are provided with track histories that capture object location, heading, velocity, and category. In all three datasets, each scenario contains its own HD Map with 3D lane and crosswalk geometry - sourced from data captured in six distinct cities. We believe these datasets will support new and existing machine learning research problems in ways that existing datasets do not. All datasets are released under the CC BY-NC-SA 4.0 license.

There are multiple scales of abstraction from which we can describe the same image, depending on whether we are focusing on fine-grained details or a more global attribute of the image. In brain mapping, learning to automatically parse images to build representations of both small-scale features (e.g., the presence of cells or blood vessels) and global properties of an image (e.g., which brain region the image comes from) is a crucial and open challenge. However, most existing datasets and benchmarks for neuroanatomy consider only a single downstream task at a time. To bridge this gap, we introduce a new dataset, annotations, and multiple downstream tasks that provide diverse ways to readout information about brain structure and architecture from the same image. Our multi-task neuroimaging benchmark (MTNeuro) is built on volumetric, micrometer-resolution X-ray microtomography images spanning a large thalamocortical section of mouse brain, encompassing multiple cortical and subcortical regions. We generated a number of different prediction challenges and evaluated several supervised and self-supervised models for brain-region prediction and pixel-level semantic segmentation of microstructures. Our experiments not only highlight the rich heterogeneity of this dataset, but also provide insights into how self-supervised approaches can be used to learn representations that capture multiple attributes of a single image and perform well on a variety of downstream tasks. Datasets, code, and pre-trained baseline models are provided at: https://mtneuro.github.io/ .

Aligning users across networks using graph representation learning has been found effective where the alignment is accomplished in a low-dimensional embedding space. Yet, achieving highly precise alignment is still challenging, especially when nodes with long-range connectivity to the labeled anchors are encountered. To alleviate this limitation, we purposefully designed WL-Align which adopts a regularized representation learning framework to learn distinctive node representations. It extends the Weisfeiler-Lehman Isormorphism Test and learns the alignment in alternating phases of "across-network Weisfeiler-Lehman relabeling" and "proximity-preserving representation learning". The across-network Weisfeiler-Lehman relabeling is achieved through iterating the anchor-based label propagation and a similarity-based hashing to exploit the known anchors' connectivity to different nodes in an efficient and robust manner. The representation learning module preserves the second-order proximity within individual networks and is regularized by the across-network Weisfeiler-Lehman hash labels. Extensive experiments on real-world and synthetic datasets have demonstrated that our proposed WL-Align outperforms the state-of-the-art methods, achieving significant performance improvements in the "exact matching" scenario. Data and code of WL-Align are available at https://github.com/ChenPengGang/WLAlignCode.

We investigate how humans perform the task of dubbing video content from one language into another, leveraging a novel corpus of 319.57 hours of video from 54 professionally produced titles. This is the first such large-scale study we are aware of. The results challenge a number of assumptions commonly made in both qualitative literature on human dubbing and machine-learning literature on automatic dubbing, arguing for the importance of vocal naturalness and translation quality over commonly emphasized isometric (character length) and lip-sync constraints, and for a more qualified view of the importance of isochronic (timing) constraints. We also find substantial influence of the source-side audio on human dubs through channels other than the words of the translation, pointing to the need for research on ways to preserve speech characteristics, as well as semantic transfer such as emphasis/emotion, in automatic dubbing systems.

This work presents a detailed linguistic analysis into why larger Transformer-based pre-trained language models with more parameters and lower perplexity nonetheless yield surprisal estimates that are less predictive of human reading times. First, regression analyses show a strictly monotonic, positive log-linear relationship between perplexity and fit to reading times for the more recently released five GPT-Neo variants and eight OPT variants on two separate datasets, replicating earlier results limited to just GPT-2 (Oh et al., 2022). Subsequently, analysis of residual errors reveals a systematic deviation of the larger variants, such as underpredicting reading times of named entities and making compensatory overpredictions for reading times of function words such as modals and conjunctions. These results suggest that the propensity of larger Transformer-based models to 'memorize' sequences during training makes their surprisal estimates diverge from humanlike expectations, which warrants caution in using pre-trained language models to study human language processing.