Counting is a fundamental example of generalization, whether viewed through the mathematical lens of Peano's axioms defining the natural numbers or the cognitive science literature for children learning to count. The argument holds for both cases that learning to count means learning to count infinitely. While few papers have tried to distill transformer "reasoning" to the simplest case of counting, investigating length generalization does occur throughout the literature. In the "train short, test long" paradigm of NLP, length refers to the training sentence length. In formal language recognition, length refers to the input sequence length, or the maximum stack size induced by a pushdown automata. In general problem solving, length refers to the number of hops in a deductive reasoning chain or the recursion depth. For all cases, counting is central to task success. And crucially, generalizing counting inductively is central to success on OOD instances. This work provides extensive empirical results on training language models to count. We experiment with architectures ranging from RNNs, Transformers, State-Space Models and RWKV. We present carefully-designed task formats, auxiliary tasks and positional embeddings to avoid limitations in generalization with OOD-position and OOD-vocabulary. We find that while traditional RNNs trivially achieve inductive counting, Transformers have to rely on positional embeddings to count out-of-domain. As counting is the basis for many arguments concerning the expressivity of Transformers, our finding calls for the community to reexamine the application scope of primitive functions defined in formal characterizations. Finally, modern RNNs also largely underperform traditional RNNs in generalizing counting inductively. We discuss how design choices that enable parallelized training of modern RNNs cause them to lose merits of a recurrent nature.
相關內容
Argumentation is a formalism allowing to reason with contradictory information by modeling arguments and their interactions. There are now an increasing number of gradual semantics and impact measures that have emerged to facilitate the interpretation of their outcomes. An impact measure assesses, for each argument, the impact of other arguments on its score. In this paper, we refine an existing impact measure from Delobelle and Villata and introduce a new impact measure rooted in Shapley values. We introduce several principles to evaluate those two impact measures w.r.t. some well-known gradual semantics. This comprehensive analysis provides deeper insights into their functionality and desirability.
To improve the perception of hierarchical structures in data sets, several color map generation algorithms have been proposed to take this structure into account. But the design of hierarchical color maps elicits different requirements to those of color maps for tabular data. Within this paper, we make an initial effort to put design rules from the color map literature into the context of hierarchical color maps. We investigate the impact of several design decisions and provide recommendations for various analysis scenarios. Thus, we lay the foundation for objective quality criteria to evaluate hierarchical color maps.
In the realm of autonomous driving, accurate 3D perception is the foundation. However, developing such models relies on extensive human annotations -- a process that is both costly and labor-intensive. To address this challenge from a data representation learning perspective, we introduce SuperFlow, a novel framework designed to harness consecutive LiDAR-camera pairs for establishing spatiotemporal pretraining objectives. SuperFlow stands out by integrating two key designs: 1) a dense-to-sparse consistency regularization, which promotes insensitivity to point cloud density variations during feature learning, and 2) a flow-based contrastive learning module, carefully crafted to extract meaningful temporal cues from readily available sensor calibrations. To further boost learning efficiency, we incorporate a plug-and-play view consistency module that enhances the alignment of the knowledge distilled from camera views. Extensive comparative and ablation studies across 11 heterogeneous LiDAR datasets validate our effectiveness and superiority. Additionally, we observe several interesting emerging properties by scaling up the 2D and 3D backbones during pretraining, shedding light on the future research of 3D foundation models for LiDAR-based perception.
MLsub is a minimal language with a type system combining subtyping and parametric polymorphism and a type inference algorithm which infers compact principal types. Simple-sub is an alternative inference algorithm which can be implemented efficiently and is easier to understand. MLsub supports explicitly typed records which are not extensible. Here we extend Simple-sub with extensible records, meaning that we can add new fields to a previously defined record. For this we add row variables to our type language and extend the type constraint solving method of our type inference algorithm accordingly, keeping the decidability of type inference.
Linear arrangements of graphs are a well-known type of graph labeling and are found in many important computational problems, such as the Minimum Linear Arrangement Problem ($\texttt{minLA}$). A linear arrangement is usually defined as a permutation of the $n$ vertices of a graph. An intuitive geometric setting is that of vertices lying on consecutive integer positions in the real line, starting at 1; edges are often drawn as semicircles above the real line. In this paper we study the Maximum Linear Arrangement problem ($\texttt{MaxLA}$), the maximization variant of $\texttt{minLA}$. We devise a new characterization of maximum arrangements of general graphs, and prove that $\texttt{MaxLA}$ can be solved for cycle graphs in constant time, and for $k$-linear trees ($k\le2$) in time $O(n)$. We present two constrained variants of $\texttt{MaxLA}$ we call $\texttt{bipartite MaxLA}$ and $\texttt{1-thistle MaxLA}$. We prove that the former can be solved in time $O(n)$ for any bipartite graph; the latter, by an algorithm that typically runs in time $O(n^4)$ on unlabelled trees. The combination of the two variants has two promising characteristics. First, it solves $\texttt{MaxLA}$ for almost all trees consisting of a few tenths of nodes. Second, we prove that it constitutes a $3/2$-approximation algorithm for $\texttt{MaxLA}$ for trees. Furthermore, we conjecture that $\texttt{bipartite MaxLA}$ solves $\texttt{MaxLA}$ for at least $50\%$ of all free trees.
A quantum computing simulation provides the opportunity to explore the behaviors of quantum circuits, study the properties of quantum gates, and develop quantum computing algorithms. Simulating quantum circuits requires geometric time and space complexities, impacting the size of the quantum circuit that can be simulated as well as the respective time required to simulate a particular circuit. Applying the parallelism inherent in the simulation and crafting custom architectures, larger quantum circuits can be simulated. A scalable accelerator architecture is proposed to provide a high performance, highly parallel, accelerator. Among the challenges of creating a scalable architecture is managing parallelism, efficiently routing quantum state components for gate evaluation, and measurement. An example is demonstrated on an Intel Agilex field programmable gate array (FPGA).
2D-based Industrial Anomaly Detection has been widely discussed, however, multimodal industrial anomaly detection based on 3D point clouds and RGB images still has many untouched fields. Existing multimodal industrial anomaly detection methods directly concatenate the multimodal features, which leads to a strong disturbance between features and harms the detection performance. In this paper, we propose Multi-3D-Memory (M3DM), a novel multimodal anomaly detection method with hybrid fusion scheme: firstly, we design an unsupervised feature fusion with patch-wise contrastive learning to encourage the interaction of different modal features; secondly, we use a decision layer fusion with multiple memory banks to avoid loss of information and additional novelty classifiers to make the final decision. We further propose a point feature alignment operation to better align the point cloud and RGB features. Extensive experiments show that our multimodal industrial anomaly detection model outperforms the state-of-the-art (SOTA) methods on both detection and segmentation precision on MVTec-3D AD dataset. Code is available at //github.com/nomewang/M3DM.
Contrastive learning models have achieved great success in unsupervised visual representation learning, which maximize the similarities between feature representations of different views of the same image, while minimize the similarities between feature representations of views of different images. In text summarization, the output summary is a shorter form of the input document and they have similar meanings. In this paper, we propose a contrastive learning model for supervised abstractive text summarization, where we view a document, its gold summary and its model generated summaries as different views of the same mean representation and maximize the similarities between them during training. We improve over a strong sequence-to-sequence text generation model (i.e., BART) on three different summarization datasets. Human evaluation also shows that our model achieves better faithfulness ratings compared to its counterpart without contrastive objectives.
Recently, a considerable literature has grown up around the theme of Graph Convolutional Network (GCN). How to effectively leverage the rich structural information in complex graphs, such as knowledge graphs with heterogeneous types of entities and relations, is a primary open challenge in the field. Most GCN methods are either restricted to graphs with a homogeneous type of edges (e.g., citation links only), or focusing on representation learning for nodes only instead of jointly propagating and updating the embeddings of both nodes and edges for target-driven objectives. This paper addresses these limitations by proposing a novel framework, namely the Knowledge Embedding based Graph Convolutional Network (KE-GCN), which combines the power of GCNs in graph-based belief propagation and the strengths of advanced knowledge embedding (a.k.a. knowledge graph embedding) methods, and goes beyond. Our theoretical analysis shows that KE-GCN offers an elegant unification of several well-known GCN methods as specific cases, with a new perspective of graph convolution. Experimental results on benchmark datasets show the advantageous performance of KE-GCN over strong baseline methods in the tasks of knowledge graph alignment and entity classification.
Data augmentation has been widely used to improve generalizability of machine learning models. However, comparatively little work studies data augmentation for graphs. This is largely due to the complex, non-Euclidean structure of graphs, which limits possible manipulation operations. Augmentation operations commonly used in vision and language have no analogs for graphs. Our work studies graph data augmentation for graph neural networks (GNNs) in the context of improving semi-supervised node-classification. We discuss practical and theoretical motivations, considerations and strategies for graph data augmentation. Our work shows that neural edge predictors can effectively encode class-homophilic structure to promote intra-class edges and demote inter-class edges in given graph structure, and our main contribution introduces the GAug graph data augmentation framework, which leverages these insights to improve performance in GNN-based node classification via edge prediction. Extensive experiments on multiple benchmarks show that augmentation via GAug improves performance across GNN architectures and datasets.
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191