Ningtao Wang

Minjie Qiang, Mingming Zhang, Xiaoyi Bao et al.

Foundation models have established unified representations for natural language processing, yet this paradigm remains largely unexplored for tabular data. Existing methods face fundamental limitations: LLM-based approaches lack retrieval-compatible vector outputs, whereas text embedding models often fail to capture tabular structure and numerical semantics. To bridge this gap, we first introduce the Tabular Embedding Benchmark (TabBench), a comprehensive suite designed to evaluate the tabular understanding capability of embedding models. We then propose TabEmbed, the first generalist embedding model that unifies tabular classification and retrieval within a shared embedding space. By reformulating diverse tabular tasks as semantic matching problems, TabEmbed leverages large-scale contrastive learning with positive-aware hard negative mining to discern fine-grained structural and numerical nuances. Experimental results on TabBench demonstrate that TabEmbed significantly outperforms state-of-the-art text embedding models, establishing a new baseline for universal tabular representation learning. Code and datasets are publicly available at https://github.com/qiangminjie27/TabEmbed and https://huggingface.co/datasets/qiangminjie27/TabBench.

Hao Chen, Qi Zhang, Liyao Li et al.

Adapting Large Language Models (LLMs) to specialized domains typically incurs high data and computational overhead. While prior efficiency efforts have largely treated data selection and parameter-efficient fine-tuning as isolated processes, our empirical analysis suggests they may be intrinsically coupled. We posit the Strong Map Hypothesis: a sparse subset of attention heads plays a dominant role in task-specific adaptation, acting as keys that unlock specific data patterns. Building on this observation, we propose From Parameters to Data (P2D), a unified framework that leverages these task-sensitive attention heads as a dual compass for both sample mining and structural pruning. To rigorously quantify the total pipeline cost, we introduce the Alignment Efficiency Ratio (AER) metric for both selection latency and training time. Mechanistically, P2D identifies critical heads via a lightweight proxy and uses them as a functional filter to curate high-affinity data, establishing a synergistic pipeline. Empirically, by updating merely 10% of attention heads on 10% of the data, P2D achieves an 8.3 pp performance gain over strong baselines and delivers a 7.0x end-to-end time speedup. These results validate that precise parameter-data synchronization eliminates redundancy, offering a new paradigm for efficient alignment.

Jiawang Dan, Ruofan Wu, Yunpeng Liu et al.

Graph representation learning has now become the de facto standard when handling graph-structured data, with the framework of message-passing graph neural networks (MPNN) being the most prevailing algorithmic tool. Despite its popularity, the family of MPNNs suffers from several drawbacks such as transparency and expressivity. Recently, the idea of designing neural models on graphs using the theory of graph kernels has emerged as a more transparent as well as sometimes more expressive alternative to MPNNs known as kernel graph neural networks (KGNNs). Developments on KGNNs are currently a nascent field of research, leaving several challenges from algorithmic design and adaptation to other learning paradigms such as self-supervised learning. In this paper, we improve the design and learning of KGNNs. Firstly, we extend the algorithmic formulation of KGNNs by allowing a more flexible graph-level similarity definition that encompasses former proposals like random walk graph kernel, as well as providing a smoother optimization objective that alleviates the need of introducing combinatorial learning procedures. Secondly, we enhance KGNNs through the lens of self-supervision via developing a novel structure-preserving graph data augmentation method called latent graph augmentation (LGA). Finally, we perform extensive empirical evaluations to demonstrate the efficacy of our proposed mechanisms. Experimental results over benchmark datasets suggest that our proposed model achieves competitive performance that is comparable to or sometimes outperforming state-of-the-art graph representation learning frameworks with or without self-supervision on graph classification tasks. Comparisons against other previously established graph data augmentation methods verify that the proposed LGA augmentation scheme captures better semantics of graph-level invariance.

Liyao Li, Chao Ye, Wentao Ye et al.

To migrate the remarkable successes of Large Language Models (LLMs), the community has made numerous efforts to generalize them to the table reasoning tasks for the widely deployed tabular data. Despite that, in this work, by showing a probing experiment on our proposed StructQA benchmark, we postulate that even the most advanced LLMs (such as GPTs) may still fall short of coping with tabular data. More specifically, the current scheme often simply relies on serializing the tabular data, together with the meta information, then inputting them through the LLMs. We argue that the loss of structural information is the root of this shortcoming. In this work, we further propose TAMO, which bears an ideology to treat the tables as an independent modality integrated with the text tokens. The resulting model in TAMO is a multimodal framework consisting of a hypergraph neural network as the global table encoder seamlessly integrated with the mainstream LLM. Empirical results on various benchmarking datasets, including HiTab, WikiTQ, WikiSQL, FeTaQA, and StructQA, have demonstrated significant improvements on generalization with an average relative gain of 42.65%.

Kejin Liu, Junhong Lian, Xiang Ao et al.

Accurate personalized headline generation hinges on precisely capturing user interests from historical behaviors. However, existing methods neglect personalized-irrelevant click noise in entire historical clickstreams, which may lead to hallucinated headlines that deviate from genuine user preferences. In this paper, we reveal the detrimental impact of click noise on personalized generation quality through rigorous analysis in both user and news dimensions. Based on these insights, we propose a novel Personalized Headline Generation framework via Denoising Fake Interests from Implicit Feedback (PHG-DIF). PHG-DIF first employs dual-stage filtering to effectively remove clickstream noise, identified by short dwell times and abnormal click bursts, and then leverages multi-level temporal fusion to dynamically model users' evolving and multi-faceted interests for precise profiling. Moreover, we release DT-PENS, a new benchmark dataset comprising the click behavior of 1,000 carefully curated users and nearly 10,000 annotated personalized headlines with historical dwell time annotations. Extensive experiments demonstrate that PHG-DIF substantially mitigates the adverse effects of click noise and significantly improves headline quality, achieving state-of-the-art (SOTA) results on DT-PENS. Our framework implementation and dataset are available at https://github.com/liukejin-up/PHG-DIF.

Hao Chen, Haoze Li, Zhiqing Xiao et al.

Aligning general-purpose large language models (LLMs) to downstream tasks often incurs significant training adjustment costs. Prior research has explored various avenues to enhance alignment efficiency, primarily through minimal-data training or data-driven activations to identify key attention heads. However, these approaches inherently introduce data dependency, which hinders generalization and reusability. To address this issue and enhance model alignment efficiency, we propose the Attention Localization and Pruning Strategy (ALPS), an efficient algorithm that localizes the most task-sensitive attention heads and prunes by restricting attention training updates to these heads, thereby reducing alignment costs. Experimental results demonstrate that our method activates only 10% of attention parameters during fine-tuning while achieving a 2% performance improvement over baselines on three tasks. Moreover, the identified task-specific heads are transferable across datasets and mitigate knowledge forgetting. Our work and findings provide a novel perspective on efficient LLM alignment. The code is available at https://github.com/VoiceBeer/ALPS.

Yin Jin, Ningtao Wang, Ruofan Wu et al.

Imbalanced data are frequently encountered in real-world classification tasks. Previous works on imbalanced learning mostly focused on learning with a minority class of few samples. However, the notion of imbalance also applies to cases where the minority class contains abundant samples, which is usually the case for industrial applications like fraud detection in the area of financial risk management. In this paper, we take a population-level approach to imbalanced learning by proposing a new formulation called \emph{ultra-imbalanced classification} (UIC). Under UIC, loss functions behave differently even if infinite amount of training samples are available. To understand the intrinsic difficulty of UIC problems, we borrow ideas from information theory and establish a framework to compare different loss functions through the lens of statistical information. A novel learning objective termed Tunable Boosting Loss is developed which is provably resistant against data imbalance under UIC, as well as being empirically efficient verified by extensive experimental studies on both public and industrial datasets.

Mingming Zhang, Pengfei Shi, Zhiqing Xiao et al.

Predictive modeling on web-scale tabular data with billions of instances and hundreds of heterogeneous numerical features faces significant scalability challenges. These features exhibit anisotropy, heavy-tailed distributions, and non-stationarity, creating bottlenecks for models like Gradient Boosting Decision Trees and requiring laborious manual feature engineering. We introduce KMLP, a hybrid deep architecture integrating a shallow Kolmogorov-Arnold Network (KAN) front-end with a Gated Multilayer Perceptron (gMLP) backbone. The KAN front-end uses learnable activation functions to automatically model complex non-linear transformations for each feature, while the gMLP backbone captures high-order interactions. Experiments on public benchmarks and an industrial dataset with billions of samples show KMLP achieves state-of-the-art performance, with advantages over baselines like GBDTs increasing at larger scales, validating KMLP as a scalable deep learning paradigm for large-scale web tabular data.

Pengfei Shi, Wei Zhong, Xinyu Zhang et al.

Estimating the conditional average treatment effects (CATE) is very important in causal inference and has a wide range of applications across many fields. In the estimation process of CATE, the unconfoundedness assumption is typically required to ensure the identifiability of the regression problems. When estimating CATE using high-dimensional data, there have been many variable selection methods and neural network approaches based on representation learning, while these methods do not provide a way to verify whether the subset of variables after dimensionality reduction or the learned representations still satisfy the unconfoundedness assumption during the estimation process, which can lead to ineffective estimates of the treatment effects. Additionally, these methods typically use data from only the treatment or control group when estimating the regression functions for each group. This paper proposes a novel neural network approach named \textbf{CrossNet} to learn a sufficient representation for the features, based on which we then estimate the CATE, where cross indicates that in estimating the regression functions, we used data from their own group as well as cross-utilized data from another group. Numerical simulations and empirical results demonstrate that our method outperforms the competitive approaches.

Mingming Zhang, Jiahao Hu, Pengfei Shi et al.

Tabular data plays a critical role in real-world financial scenarios. Traditionally, tree models have dominated in handling tabular data. However, financial datasets in the industry often encounter some challenges, such as data heterogeneity, the predominance of numerical features and the large scale of the data, which can range from tens of millions to hundreds of millions of records. These challenges can lead to significant memory and computational issues when using tree-based models. Consequently, there is a growing need for neural network-based solutions that can outperform these models. In this paper, we introduce TKGMLP, an hybrid network for tabular data that combines shallow Kolmogorov Arnold Networks with Gated Multilayer Perceptron. This model leverages the strengths of both architectures to improve performance and scalability. We validate TKGMLP on a real-world credit scoring dataset, where it achieves state-of-the-art results and outperforms current benchmarks. Furthermore, our findings demonstrate that the model continues to improve as the dataset size increases, making it highly scalable. Additionally, we propose a novel feature encoding method for numerical data, specifically designed to address the predominance of numerical features in financial datasets. The integration of this feature encoding method within TKGMLP significantly improves prediction accuracy. This research not only advances table prediction technology but also offers a practical and effective solution for handling large-scale numerical tabular data in various industrial applications.

Zeyu Zhao, Ningtao Wang, Xing Fu et al.

Encoder-only Transformers have advanced along three axes -- architecture, data, and systems -- yielding Pareto gains in accuracy, speed, and memory efficiency. Yet these improvements have not fully transferred to Chinese, where tokenization and morphology differ markedly from English. We introduce Chinese ModernBERT, a from-scratch Chinese encoder that couples: (i) a hardware-aware 32k BPE vocabulary tailored to frequent Chinese affixes/compounds, lowering the embedding budget; (ii) whole-word masking (WWM) with a dynamic masking curriculum (30% -> 15%) to align task difficulty with training progress; (iii) a two-stage pre-training pipeline that extends the native context from 1,024 to 8,192 tokens using RoPE and alternating local/global attention; and (iv) a damped-cosine learning-rate schedule for stable long-horizon optimization. We pre-train on ~1.2T Chinese tokens from CCI3-HQ, CCI4 (Chinese), and Cosmopedia-Chinese. On CLUE, Chinese ModernBERT is competitive with strong Chinese encoders under a unified fine-tuning protocol. Under bf16 it achieves high long-sequence throughput while maintaining strong short-sequence speed, reflecting benefits from budget allocation and attention design. To probe retrieval-oriented quality, we add a small amount of open contrastive data: fine-tuning on SimCLUE (~3M pairs) improves further when adding T2Ranking (~2M), reaching 0.505 (Pearson) / 0.537 (Spearman) on the SimCLUE test set. Under this open-data setting, Chinese ModernBERT surpasses Qwen-0.6B-embedding on SimCLUE, suggesting a clear scaling path for STS with additional curated pairs. We will release tokenizer and weights to facilitate reproducible research.