This paper proposes a variable range-based interaction preference multi-objective optimization algorithm through objective decomposition (VR-MOEA-OD). VR-MOEA-OD focuses on finding the preferred solutions of decision maker (DM) and can adjust the size of the preference range based on the DM’s needs. First, the DM is asked to provide a preference direction and the number of needed solutions. Then, the algorithm searches for the DM’s preferred solutions using the multi-objective evolutionary algorithm based on objective decomposition (MOEA-OD). If unsatisfied with the results, the DM needs to provide a range parameter to adjust the size of preference range and respecify the number of desired solutions. Based on the range parameter and the change in the number of solutions required before and after the interaction, VR-MOEA-OD can modulate the size of the preference range, either enlarging or narrowing the preference range. The results indicate that VR-MOEA-OD can fully utilize preference information and satisfy the interaction information provided by the DM to achieve desired results.

Tabular data is one of the most ubiquitous sources of information worldwide, spanning a wide variety of domains. This inherent heterogeneity has slowed the development of Tabular Foundation Models (TFMs) capable of fast generalization to unseen datasets. In-Context Learning (ICL) has recently emerged as a promising solution for TFMs, enabling dynamic adaptation to new tasks without additional tuning. While many studies have attempted to re-purpose large language models for tabular ICL, they have had limited success, so recent works have focused on developing tabular-specific foundation models. In this work, we propose an approach to combine ICL-based retrieval with self supervised learning to train tabular foundation models. We also investigate the utility of real vs. synthetic data for model pre-training, and show that real data can contain useful signal not easily captured in synthetic training. Specifically, we show that incorporating real data during the pre-training phase can lead to significantly faster training and better downstream generalization to unseen data. Our resulting model, **TabDPT**, achieves strong performance on both regression (CTR23) and classification (CC18) benchmarks. Importantly, we also demonstrate that with our pre-training procedure, scaling both model and data size leads to consistent performance improvements that follow power laws. This echoes scaling laws in LLMs and other foundation models, and suggests that large-scale TFMs can be achievable. We open-source our full pipeline: inference code including trained model weights can be found [here](https://github.com/layer6ai-labs/TabDPT-inference), and the training code to reproduce experiments can be found [here](https://github.com/layer6ai-labs/TabDPT-training).

Causal effect estimation from observational data is fundamental across various applications. However, selecting an appropriate estimator from dozens of specialized methods demands substantial manual effort and domain expertise. We present CausalPFN, a single transformer that *amortizes* this workflow: trained once on a large library of simulated data-generating processes that satisfy ignorability, it infers causal effects for new observational datasets out of the box. CausalPFN combines ideas from Bayesian causal inference with the large-scale training protocol of prior-fitted networks (PFNs), learning to map raw observations directly to causal effects without any task-specific adjustment. Our approach achieves superior average performance on heterogeneous and average treatment effect estimation benchmarks (IHDP, Lalonde, ACIC). Moreover, it shows competitive performance for real-world policy making on uplift modeling tasks. CausalPFN provides calibrated uncertainty estimates to support reliable decision-making based on Bayesian principles. This ready-to-use model requires no further training or tuning and takes a step toward automated causal inference (https://github.com/vdblm/CausalPFN/).

The application of smart contracts in electric power systems is widespread. However, vulnerabilities in smart contracts can cause significant economic losses and require careful attention. Smart contracts in electric power systems have domain-specific characteristics that differ from traditional public blockchain applications. As a result, existing vulnerability detection tools cannot be directly applied to these systems. To address this challenge, we design a vulnerability detection tool called E-Guard specifically for smart contracts in electric power systems. E-Guard uses a tailored intermediate representation (IR) known as EIR, which provides control flow and data flow information more suited to the business logic of electric power systems than traditional static analysis tools. We identify and summarize three types of vulnerabilities unique to electric power systems based on expert knowledge. Experimental results show that E-Guard significantly outperforms traditional static analysis tools in detecting these three types of vulnerabilities. Additionally, the extra overhead generated by using EIR is minimal and negligible. This demonstrates that E-Guard is an effective and efficient tool for enhancing the security of smart contracts in electric power systems.

Tabular data is a pervasive modality spanning a wide range of domains, and this inherent diversity poses a considerable challenge for deep learning. Recent advancements using transformer-based in-context learning have shown promise on smaller and less complex tabular datasets, but have struggled to scale to larger and more complex ones. To address this limitation, we propose a combination of retrieval and fine-tuning: we can adapt the transformer to a local subset of the data by collecting nearest neighbours, and then perform task-specific fine-tuning with this retrieved set of neighbours in context. Using TabPFN as the base model -- currently the best tabular in-context learner -- and applying our retrieval and fine-tuning scheme on top results in what we call a locally-calibrated PFN, or LoCalPFN. We conduct extensive evaluation on 95 datasets curated by TabZilla from OpenML, upon which we establish a new state-of-the-art with LoCalPFN -- even with respect to tuned tree-based models. Notably, we show a significant boost in performance compared to the base in-context model, demonstrating the efficacy of our approach and advancing the frontier of deep learning in tabular data.

In text-video retrieval, the objective is to learn a cross-modal similarity function between a text and a video that ranks relevant text-video pairs higher than irrelevant pairs. However, videos inherently express a much wider gamut of information than texts. Instead, texts often capture sub-regions of entire videos and are most semantically similar to certain frames within videos. Therefore, for a given text, a retrieval model should focus on the text's most semantically similar video sub-regions to make a more relevant comparison. Yet, most existing works aggregate entire videos with-out directly considering text. Common text-agnostic ag-gregations schemes include mean-pooling or self-attention over the frames, but these are likely to encode misleading vi-sual information not described in the given text. To address this, we propose a cross-modal attention model called X-Pool that reasons between a text and the frames of a video. Our core mechanism is a scaled dot product attention for a text to attend to its most semantically similar frames. We then generate an aggregated video representation conditioned on the text's attention weights over the frames. We evaluate our method on three benchmark datasets of MSR-VTT, MSVD and LSMDC, achieving new state-of-the-art re-sults by up to 12% in relative improvement in Recall@ 1. Our findings thereby highlight the importance of joint text-video reasoning to extract important visual cues according to text. Full code and demo can be found at: layer6ai-labs.github.iolxpooll.

With outbreak of the COVID-19 pandemic, contact tracing has become an important problem. It has been proven that maintaining social distance and isolating affected people are highly beneficial for curbing the spread of COVID-19, which all depend on identifying people’s trajectories. However, the current interview-based approach is costly, and the existing mobile app-based schemes rely on complete and accurate data. In this paper, we propose a transformer encoder-based approach with spatial position embedding extracted using a graph Combinatorial Laplacian matrix to interpolate incomplete human trajectories. To model human trajectory, we propose a graphical embedded module to extract spatial features based on predefined location clusters. The incomplete trajectory sequences are first preprocessed into matrices and then used to train a deep transformer encoder network for trajectory completion. Our experiments using a real world Bluetooth Low Energy (BLE) dataset validate the efficacy of our proposed approach, which outperforms several baseline methods.

Machine learning (ML) has been extensively applied to model geohazards, yielding tremendous success. However, researchers and practitioners still face challenges in enhancing the reliability of ML models. In the present study, a systematic framework combining k-fold cross-validation (CV), metaheuristics (MHs), support vector regression (SVR), and Friedman and Nemenyi tests was proposed to improve the reliability and performance of geohazard modeling. The average normalized mean square error (NMSE) from k-fold CV sets was adopted as the fitness metric. Twenty of the most well-established MHs and the most recent MHs were adopted to tune the hyperparameters of SVR and were evaluated through nonparametric Friedman and post hoc Nemenyi tests to identify significant differences. Observations from a typical reservoir landslide were selected as a benchmark dataset, and the accuracy, robustness, computational time, and convergence speed of the MHs were compared. Significant performance differences among the twenty MHs were identified by Friedman and post hoc Nemenyi tests of the mean absolute error (MAE), root mean squared error (RMSE), Kling–Gupta efficiency (KGE), and computational time, with p values lower than 0.05. The comparison of results demonstrated that the multiverse optimizer (MVO) is among the highest-performing, most stable, and computationally efficient algorithms, providing superior performance to other methods, with nearly optimum values of the correlation coefficient (R), a low MAE (23.5086 versus 23.9360), a low mean RMSE (48.6946 versus 50.1882), and a high mean KGE (0.9803 versus 0.9893) in predicting the displacement of the Shuping landslide. This paper considerably enriches the literature regarding hyperparameter optimization algorithms and the enhancement of their reliability. In addition, Friedman and post hoc Nemenyi tests have the potential for evaluating and comparing various ML-based geohazard models.

Localizing actions in video is a core task in computer vision. The weakly supervised temporal localization problem investigates whether this task can be adequately solved with only video-level labels, significantly reducing the amount of expensive and error-prone annotation that is required. A common approach is to train a frame-level classifier where frames with the highest class probability are selected to make a video-level prediction. Frame-level activations are then used for localization. However, the absence of frame-level annotations cause the classifier to impart class bias on every frame. To address this, we propose the Action Selection Learning (ASL) approach to capture the general concept of action, a property we refer to as "actionness". Under ASL, the model is trained with a novel class-agnostic task to predict which frames will be selected by the classifier. Empirically, we show that ASL outperforms leading baselines on two popular bench-marks THUMOS-14 and ActivityNet-1.2, with 10.3% and 5.7% relative improvement respectively. We further analyze the properties of ASL and demonstrate the importance of actionness. Full code for this work is available here: https://github.com/layer6ai-labs/ASL.

Despite recent progress in computer vision, image retrieval remains a challenging open problem. Numerous variations such as view angle, lighting and occlusion make it difficult to design models that are both robust and efficient. Many leading methods traverse the nearest neighbor graph to exploit higher order neighbor information and uncover the highly complex underlying manifold. In this work we propose a different approach where we leverage graph convolutional networks to directly encode neighbor information into image descriptors. We further leverage ideas from clustering and manifold learning, and introduce an unsupervised loss based on pairwise separation of image similarities. Empirically, we demonstrate that our model is able to successfully learn a new descriptor space that significantly improves retrieval accuracy, while still allowing efficient inner product inference. Experiments on five public benchmarks show highly competitive performance with up to 24\% relative improvement in mAP over leading baselines. Full code for this work is available here: https://github.com/layer6ai-labs/GSS.

We present an automatic approach to detecting symmetry relations for general concurrent models. Despite the success of symmetry reduction in mitigating state explosion problem, one essential step towards its soundness and effectiveness, i.e., how to discover sufficient symmetries with least human efforts, is often either overlooked or oversimplified. In this work, we show how a concurrent model can be viewed as a constraint satisfaction problem (CSP), and present an algorithm capable of detecting symmetries arising from the CSP which induce automorphisms of the model. To the best of our knowledge, our method is the first approach that can automatically detect both process and data symmetries as demonstrated via a number of systems.

Many species of Gram-negative bacteria are pathogenic bacteria that can cause disease in a host organism. This pathogenic capability is usually associated with certain components in Gram-negative cells, so it is highly desirable to develop an effective method to predict the Gram-negative bacterial protein subcellular locations. Reflecting the wide applications of neural networks in this field, we design seven different training functions based on Elman networks, and use a genetic algorithm to select the proper networks for an ensemble. Experimental results show that the neural networks ensemble has a dominant advantage in performance.