PyTorchKR Papers

최신 논문

총 139,333편
태그·기간 등 상세 필터

카테고리 탐색

arXiv:2505.09246공식 코드 · Papers with Code

Focus, Merge, Rank: Improved Question Answering Based on Semi-structured Knowledge Bases

In many real-world settings, machine learning models and interactive systems have access to both structured knowledge, e.g., knowledge graphs or tables, and unstructured content, e.g., natural language documents. However, most rely on either. Semi-Structured Knowledge Bases (SKBs) bridge this gap by linking unstructured content to nodes within structured data, thereby enabling new strategies for knowledge access and use. In this work, we present FocusedRetriever, a modular SKB-based framework for multi-hop question answering. It integrates components (VSS-based entity search, LLM-based generation of Cypher queries and pairwise re-ranking) in a way that enables it to outperform state-of-the-art methods across all three STaRK benchmark test sets, covering diverse domains and multiple performance metrics. The average first-hit rate exceeds that of the second-best method by 25.7%. FocusedRetriever leverages (1) the capacity of Large Language Models (LLMs) to extract relational facts and entity attributes from unstructured text, (2) node set joins to filter answer candidates based on these extracted triplets and constraints, (3) vector similarity search to retrieve and rank relevant unstructured content, and (4) the contextual capabilities of LLMs to finally rank the top-k answers. For generality, we only incorporate base LLMs in FocusedRetriever in our evaluation. However, our analysis of intermediate results highlights several opportunities for further upgrades including finetuning. The source code is publicly available at https://github.com/kramerlab/FocusedRetriever .

arXiv:2505.09203공식 코드 · Papers with Code

InvDesFlow-AL: Active Learning-based Workflow for Inverse Design of Functional Materials

Developing inverse design methods for functional materials with specific properties is critical to advancing fields like renewable energy, catalysis, energy storage, and carbon capture. Generative models based on diffusion principles can directly produce new materials that meet performance constraints, thereby significantly accelerating the material design process. However, existing methods for generating and predicting crystal structures often remain limited by low success rates. In this work, we propose a novel inverse material design generative framework called InvDesFlow-AL, which is based on active learning strategies. This framework can iteratively optimize the material generation process to gradually guide it towards desired performance characteristics. In terms of crystal structure prediction, the InvDesFlow-AL model achieves an RMSE of 0.0423 {\AA}, representing an 32.96% improvement in performance compared to exsisting generative models. Additionally, InvDesFlow-AL has been successfully validated in the design of low-formation-energy and low-Ehull materials. It can systematically generate materials with progressively lower formation energies while continuously expanding the exploration across diverse chemical spaces. These results fully demonstrate the effectiveness of the proposed active learning-driven generative model in accelerating material discovery and inverse design. To further prove the effectiveness of this method, we took the search for BCS superconductors under ambient pressure as an example explored by InvDesFlow-AL. As a result, we successfully identified Li\(_2\)AuH\(_6\) as a conventional BCS superconductor with an ultra-high transition temperature of 140 K. This discovery provides strong empirical support for the application of inverse design in materials science.

arXiv:2505.09168공식 코드 · Papers with Code

DRRNet: Macro-Micro Feature Fusion and Dual Reverse Refinement for Camouflaged Object Detection

The core challenge in Camouflage Object Detection (COD) lies in the indistinguishable similarity between targets and backgrounds in terms of color, texture, and shape. This causes existing methods to either lose edge details (such as hair-like fine structures) due to over-reliance on global semantic information or be disturbed by similar backgrounds (such as vegetation patterns) when relying solely on local features. We propose DRRNet, a four-stage architecture characterized by a "context-detail-fusion-refinement" pipeline to address these issues. Specifically, we introduce an Omni-Context Feature Extraction Module to capture global camouflage patterns and a Local Detail Extraction Module to supplement microstructural information for the full-scene context module. We then design a module for forming dual representations of scene understanding and structural awareness, which fuses panoramic features and local features across various scales. In the decoder, we also introduce a reverse refinement module that leverages spatial edge priors and frequency-domain noise suppression to perform a two-stage inverse refinement of the output. By applying two successive rounds of inverse refinement, the model effectively suppresses background interference and enhances the continuity of object boundaries. Experimental results demonstrate that DRRNet significantly outperforms state-of-the-art methods on benchmark datasets. Our code is available at https://github.com/jerrySunning/DRRNet.

arXiv:2505.09114공식 코드 · Papers with Code

Beyond the Known: Decision Making with Counterfactual Reasoning Decision Transformer

Decision Transformers (DT) play a crucial role in modern reinforcement learning, leveraging offline datasets to achieve impressive results across various domains. However, DT requires high-quality, comprehensive data to perform optimally. In real-world applications, the lack of training data and the scarcity of optimal behaviours make training on offline datasets challenging, as suboptimal data can hinder performance. To address this, we propose the Counterfactual Reasoning Decision Transformer (CRDT), a novel framework inspired by counterfactual reasoning. CRDT enhances DT ability to reason beyond known data by generating and utilizing counterfactual experiences, enabling improved decision-making in unseen scenarios. Experiments across Atari and D4RL benchmarks, including scenarios with limited data and altered dynamics, demonstrate that CRDT outperforms conventional DT approaches. Additionally, reasoning counterfactually allows the DT agent to obtain stitching abilities, combining suboptimal trajectories, without architectural modifications. These results highlight the potential of counterfactual reasoning to enhance reinforcement learning agents' performance and generalization capabilities.

arXiv:2505.09062공식 코드 · Papers with Code

Variational Prefix Tuning for Diverse and Accurate Code Summarization Using Pre-trained Language Models

Recent advancements in source code summarization have leveraged transformer-based pre-trained models, including Large Language Models of Code (LLMCs), to automate and improve the generation of code summaries. However, existing methods often focus on generating a single high-quality summary for a given source code, neglecting scenarios where the generated summary might be inadequate and alternative options are needed. In this paper, we introduce Variational Prefix Tuning (VPT), a novel approach that enhances pre-trained models' ability to generate diverse yet accurate sets of summaries, allowing the user to choose the most suitable one for the given source code. Our method integrates a Conditional Variational Autoencoder (CVAE) framework as a modular component into pre-trained models, enabling us to model the distribution of observed target summaries and sample continuous embeddings to be used as prefixes to steer the generation of diverse outputs during decoding. Importantly, we construct our method in a parameter-efficient manner, eliminating the need for expensive model retraining, especially when using LLMCs. Furthermore, we employ a bi-criteria reranking method to select a subset of generated summaries, optimizing both the diversity and the accuracy of the options presented to users. We present extensive experimental evaluations using widely used datasets and current state-of-the-art pre-trained code summarization models to demonstrate the effectiveness of our approach and its adaptability across models.

arXiv:2505.08053공식 코드 · Papers with Code

Preventing SARS-CoV-2 superspreading events with antiviral intranasal sprays

Superspreading events are known to disproportionally contribute to onwards transmission of epidemic and pandemic viruses. Preventing infections at a small number of high-transmission settings is therefore an attractive public health goal. Here, we use deterministic and stochastic mathematical modelling to quantify the impact of intranasal sprays in containing outbreaks at a known superspreading event (the 2020 SARS-CoV-2 outbreak at the Diamond Princess cruise ship) and a conference event that led to extensive transmission. We find that in the Diamond Princess cruise ship case study, there exists a 7-14-day window of opportunity for widespread prophylactic spray usage to significantly impact the number of infections averted. Given an immediate response to a known SARS-CoV-2 outbreak, alongside testing and social distancing measures, prophylactic efficacy and coverage greater than 65% could reduce the average number of infections by over 90%. In the conference case study, in the absence of additional public health interventions, analyses suggest much higher prophylactic efficacies and coverages are required to achieve a similar outcome. However, prophylactic use can half an individual's probability of being infected, and significantly reduce the probability of developing a severe infection. These results suggest that at a known potential superspreading event, early use of intranasal sprays can complement quarantining measures and significantly suppress a SARS-CoV-2 outbreak, even at suboptimal coverage. At a potential superspreading event of short duration, intranasal sprays can reduce individuals' risk of infection, but cannot prevent all infections or onwards community transmission.

arXiv:2505.07139공식 코드 · Papers with Code

Impact of SMILES Notational Inconsistencies on Chemical Language Model Performance

Chemical language models (CLMs), inspired by natural language processing (NLP), have recently emerged as powerful tools for various cheminformatics tasks such as molecular property prediction and molecule generation. The simplified molecular input line entry system (SMILES) is commonly employed in CLMs for representing molecular structures. However, despite attempts at standardization through Canonical SMILES, significant representational inconsistencies remain due to variations in canonicalization methods (grammatical inconsistencies) and incomplete stereochemical annotations (stereochemical inconsistencies). This study systematically investigates the prevalence and impact of these inconsistencies. Our literature review reveals that nearly half (45.4%) of the studies employing CLMs omit explicit mention of SMILES canonicalization, potentially impairing reproducibility. Through quantitative analysis of publicly available datasets, we observed substantial variability in SMILES representations, particularly highlighting significant gaps in stereochemical information approximately 50% of enantiomers and 30% of geometric isomers lacked complete annotations. Empirical evaluations using an encoder-decoder CLM demonstrated that representational variations significantly affect latent molecular representations, notably reducing translation accuracy. Interestingly, these variations minimally impacted downstream property prediction tasks, likely due to robust feature selection driven by label information. Furthermore, explicit manipulation of stereochemical annotations confirmed their crucial role in accurate cyclic structure reconstruction.

arXiv:2505.06950공식 코드 · Papers with Code

Copula Analysis of Risk: A Multivariate Risk Analysis for VaR and CoVaR using Copulas and DCC-GARCH

A multivariate risk analysis for VaR and CVaR using different copula families is performed on historical financial time series fitted with DCC-GARCH models. A theoretical background is provided alongside a comparison of goodness-of-fit across different copula families to estimate the validity and effectiveness of approaches discussed.

arXiv:2505.06784공식 코드 · Papers with Code

Hillclimb-Causal Inference: A Data-Driven Approach to Identify Causal Pathways Among Parental Behaviors, Genetic Risk, and Externalizing Behaviors in Children

Motivation: Externalizing behaviors in children, such as aggression, hyperactivity, and defiance, are influenced by complex interplays between genetic predispositions and environmental factors, particularly parental behaviors. Unraveling these intricate causal relationships can benefit from the use of robust data-driven methods. Methods: We developed a method called Hillclimb-Causal Inference, a causal discovery approach that integrates the Hill Climb Search algorithm with a customized Linear Gaussian Bayesian Information Criterion (BIC). This method was applied to data from the Adolescent Brain Cognitive Development (ABCD) Study, which included parental behavior assessments, children's genotypes, and externalizing behavior measures. We performed dimensionality reduction to address multicollinearity among parental behaviors and assessed children's genetic risk for externalizing disorders using polygenic risk scores (PRS), which were computed based on GWAS summary statistics from independent cohorts. Once the causal pathways were identified, we employed structural equation modeling (SEM) to quantify the relationships within the model. Results: We identified prominent causal pathways linking parental behaviors to children's externalizing outcomes. Parental alcohol misuse and broader behavioral issues exhibited notably stronger direct effects (0.33 and 0.20, respectively) compared to children's polygenic risk scores (0.07). Moreover, when considering both direct and indirect paths, parental substance misuse (alcohol, drug, and tobacco) collectively resulted in a total effect exceeding 1.1 on externalizing behaviors. Bootstrap and sensitivity analyses further validated the robustness of these findings.

arXiv:2505.00887공식 코드 · Papers with Code

Rethinking Time Encoding via Learnable Transformation Functions

Effectively modeling time information and incorporating it into applications or models involving chronologically occurring events is crucial. Real-world scenarios often involve diverse and complex time patterns, which pose significant challenges for time encoding methods. While previous methods focus on capturing time patterns, many rely on specific inductive biases, such as using trigonometric functions to model periodicity. This narrow focus on single-pattern modeling makes them less effective in handling the diversity and complexities of real-world time patterns. In this paper, we investigate to improve the existing commonly used time encoding methods and introduce Learnable Transformation-based Generalized Time Encoding (LeTE). We propose using deep function learning techniques to parameterize non-linear transformations in time encoding, making them learnable and capable of modeling generalized time patterns, including diverse and complex temporal dynamics. By enabling learnable transformations, LeTE encompasses previous methods as specific cases and allows seamless integration into a wide range of tasks. Through extensive experiments across diverse domains, we demonstrate the versatility and effectiveness of LeTE.

arXiv:2504.05172공식 코드 · Papers with Code

Attention-Based Multiscale Temporal Fusion Network for Uncertain-Mode Fault Diagnosis in Multimode Processes

Fault diagnosis in multimode processes plays a critical role in ensuring the safe operation of industrial systems across multiple modes. It faces a great challenge yet to be addressed - that is, the significant distributional differences among monitoring data from multiple modes make it difficult for the models to extract shared feature representations related to system health conditions. In response to this problem, this paper introduces a novel method called attention-based multiscale temporal fusion network. The multiscale depthwise convolution and gated recurrent unit are employed to extract multiscale contextual local features and long-short-term features. Instance normalization is applied to suppress mode-specific information. Furthermore, a temporal attention mechanism is designed to focus on critical time points with higher cross-mode shared information, thereby enhancing the accuracy of fault diagnosis. The proposed model is applied to Tennessee Eastman process dataset and three-phase flow facility dataset. The experiments demonstrate that the proposed model achieves superior diagnostic performance and maintains a small model size. The source code will be available on GitHub at https://github.com/GuangqiangLi/AMTFNet.

arXiv:2504.04740공식 코드 · Papers with Code

Enhancing Compositional Reasoning in Vision-Language Models with Synthetic Preference Data

Compositionality, or correctly recognizing scenes as compositions of atomic visual concepts, remains difficult for multimodal large language models (MLLMs). Even state of the art MLLMs such as GPT-4o can make mistakes in distinguishing compositions like "dog chasing cat" vs "cat chasing dog". While on Winoground, a benchmark for measuring such reasoning, MLLMs have made significant progress, they are still far from a human's performance. We show that compositional reasoning in these models can be improved by elucidating such concepts via data, where a model is trained to prefer the correct caption for an image over a close but incorrect one. We introduce SCRAMBLe: Synthetic Compositional Reasoning Augmentation of MLLMs with Binary preference Learning, an approach for preference tuning open-weight MLLMs on synthetic preference data generated in a fully automated manner from existing image-caption data. SCRAMBLe holistically improves these MLLMs' compositional reasoning capabilities which we can see through significant improvements across multiple vision language compositionality benchmarks, as well as smaller but significant improvements on general question answering tasks. As a sneak peek, SCRAMBLe tuned Molmo-7B model improves on Winoground from 49.5% to 54.8% (best reported to date), while improving by ~1% on more general visual question answering tasks. Code for SCRAMBLe along with tuned models and our synthetic training dataset is available at https://github.com/samarth4149/SCRAMBLe.

arXiv:2502.01391공식 코드 · Papers with Code

Learning Traffic Anomalies from Generative Models on Real-Time Observations

Accurate detection of traffic anomalies is crucial for effective urban traffic management and congestion mitigation. We use the Spatiotemporal Generative Adversarial Network (STGAN) framework combining Graph Neural Networks and Long Short-Term Memory networks to capture complex spatial and temporal dependencies in traffic data. We apply STGAN to real-time, minute-by-minute observations from 42 traffic cameras across Gothenburg, Sweden, collected over several months in 2020. The images are processed to compute a flow metric representing vehicle density, which serves as input for the model. Training is conducted on data from April to November 2020, and validation is performed on a separate dataset from November 14 to 23, 2020. Our results demonstrate that the model effectively detects traffic anomalies with high precision and low false positive rates. The detected anomalies include camera signal interruptions, visual artifacts, and extreme weather conditions affecting traffic flow.

arXiv:2505.08142공식 코드 · Papers with Code

Highly Undersampled MRI Reconstruction via a Single Posterior Sampling of Diffusion Models

Incoherent k-space under-sampling and deep learning-based reconstruction methods have shown great success in accelerating MRI. However, the performance of most previous methods will degrade dramatically under high acceleration factors, e.g., 8$\times$ or higher. Recently, denoising diffusion models (DM) have demonstrated promising results in solving this issue; however, one major drawback of the DM methods is the long inference time due to a dramatic number of iterative reverse posterior sampling steps. In this work, a Single Step Diffusion Model-based reconstruction framework, namely SSDM-MRI, is proposed for restoring MRI images from highly undersampled k-space. The proposed method achieves one-step reconstruction by first training a conditional DM and then iteratively distilling this model. Comprehensive experiments were conducted on both publicly available fastMRI images and an in-house multi-echo GRE (QSM) subject. Overall, the results showed that SSDM-MRI outperformed other methods in terms of numerical metrics (PSNR and SSIM), qualitative error maps, image fine details, and latent susceptibility information hidden in MRI phase images. In addition, the reconstruction time for a 320*320 brain slice of SSDM-MRI is only 0.45 second, which is only comparable to that of a simple U-net, making it a highly effective solution for MRI reconstruction tasks.

arXiv:2505.07913공식 코드 · Papers with Code

Unequal Journeys to Food Markets: Continental-Scale Evidence from Open Data in Africa

Food market accessibility is a critical yet underexplored dimension of food systems, particularly in low- and middle-income countries. Here, we present a continent-wide assessment of spatial food market accessibility in Africa, integrating open geospatial data from OpenStreetMap and the World Food Programme. We compare three complementary metrics: travel time to the nearest market, market availability within a 30-minute threshold, and an entropy-based measure of spatial distribution, to quantify accessibility across diverse settings. Our analysis reveals pronounced disparities: rural and economically disadvantaged populations face substantially higher travel times, limited market reach, and less spatial redundancy. These accessibility patterns align with socioeconomic stratification, as measured by the Relative Wealth Index, and moderately correlate with food insecurity levels, assessed using the Integrated Food Security Phase Classification. Overall, results suggest that access to food markets plays a relevant role in shaping food security outcomes and reflects broader geographic and economic inequalities. This framework provides a scalable, data-driven approach for identifying underserved regions and supporting equitable infrastructure planning and policy design across diverse African contexts.

arXiv:2505.07598공식 코드 · Papers with Code

Wireless Link Scheduling with State-Augmented Graph Neural Networks

We consider the problem of optimal link scheduling in large-scale wireless ad hoc networks. We specifically aim for the maximum long-term average performance, subject to a minimum transmission requirement for each link to ensure fairness. With a graph structure utilized to represent the conflicts of links, we formulate a constrained optimization problem to learn the scheduling policy, which is parameterized with a graph neural network (GNN). To address the challenge of long-term performance, we use the state-augmentation technique. In particular, by augmenting the Lagrangian dual variables as dynamic inputs to the scheduling policy, the GNN can be trained to gradually adapt the scheduling decisions to achieve the minimum transmission requirements. We verify the efficacy of our proposed policy through numerical simulations and compare its performance with several baselines in various network settings.

arXiv:2505.06127공식 코드 · Papers with Code

FastDup: a scalable duplicate marking tool using speculation-and-test mechanism

Duplicate marking is a critical preprocessing step in gene sequence analysis to flag redundant reads arising from polymerase chain reaction(PCR) amplification and sequencing artifacts. Although Picard MarkDuplicates is widely recognized as the gold-standard tool, its single-threaded implementation and reliance on global sorting result in significant computational and resource overhead, limiting its efficiency on large-scale datasets. Here, we introduce FastDup: a high-performance, scalable solution that follows the speculation-and-test mechanism. FastDup achieves up to 20x throughput speedup and guarantees 100\% identical output compared to Picard MarkDuplicates. FastDup is a C++ program available from GitHub (https://github.com/zzhofict/FastDup.git) under the MIT license.

arXiv:2505.08365공식 코드 · Papers with Code

Closed-Form Information Capacity of Canonical Signaling Models

We employ a unified framework for computing the information capacity of biological signaling systems using Fisher Information. By deriving closed-form or easily computable information capacity formulas, we quantify how well different signaling models, including binomial, multinomial, Poisson, Gaussian, and Gamma distributions, can discriminate among input signals. These expressions clarify how key features such as signal range, noise scaling, pathway length, and receivers' diversity shape the theoretical limits of sensing. In particular, we show how signal-to-noise ratio and fold-change sensitivity arise naturally within the Fisher formalism, and how signal degradation in cascades imposes linear information loss. Our results provide intuitive, analytically grounded tools to benchmark and guide the analysis of real signaling systems, without requiring computationally expensive mutual information estimation. While motivated by cellular communication, the framework generalizes to any system where noisy input-output relationships constrain transmission fidelity, including synthetic biology, sensor networks, and engineered communication channels.

arXiv:2505.07117공식 코드 · Papers with Code

OPTIKS: Optimized Gradient Properties Through Timing in K-Space

A customizable method (OPTIKS) for designing fast trajectory-constrained gradient waveforms with optimized time domain properties was developed. Given a specified multidimensional k-space trajectory, the method optimizes traversal speed (and therefore timing) with position along the trajectory. OPTIKS facilitates optimization of objectives dependent on the time domain gradient waveform and the arc-length domain k-space speed. OPTIKS is applied to design waveforms which limit peripheral nerve stimulation (PNS), minimize mechanical resonance excitation, and reduce acoustic noise. A variety of trajectory examples are presented including spirals, circular echo-planar-imaging, and rosettes. Design performance is evaluated based on duration, standardized PNS models, field measurements, gradient coil back-EMF measurements, and calibrated acoustic measurements. We show reductions in back-EMF of up to 94% and field oscillations up to 91.1%, acoustic noise decreases of up to 9.22 dB, and with efficient use of PNS models speed increases of up to 11.4%. The design method implementation is made available as an open source Python package through GitHub.

arXiv:2505.07053공식 코드 · Papers with Code

The Dynamics of Inducible Genetic Circuits

Genes are connected in complex networks of interactions where often the product of one gene is a transcription factor that alters the expression of another. Many of these networks are based on a few fundamental motifs leading to switches and oscillators of various kinds. And yet, there is more to the story than which transcription factors control these various circuits. These transcription factors are often themselves under the control of effector molecules that bind them and alter their level of activity. Traditionally, much beautiful work has shown how to think about the stability of the different states achieved by these fundamental regulatory architectures by examining how parameters such as transcription rates, degradation rates and dissociation constants tune the circuit, giving rise to behavior such as bistability. However, such studies explore dynamics without asking how these quantities are altered in real time in living cells as opposed to at the fingertips of the synthetic biologist's pipette or on the computational biologist's computer screen. In this paper, we make a departure from the conventional dynamical systems view of these regulatory motifs by using statistical mechanical models to focus on endogenous signaling knobs such as effector concentrations rather than on the convenient but more experimentally remote knobs such as dissociation constants, transcription rates and degradation rates that are often considered. We also contrast the traditional use of Hill functions to describe transcription factor binding with more detailed thermodynamic models. This approach provides insights into how biological parameters are tuned to control the stability of regulatory motifs in living cells, sometimes revealing quite a different picture than is found by using Hill functions and tuning circuit parameters by hand.