arXiv subjects

Wentao Cui

Publications and source records attributed to Wentao Cui.

9 recordsLinked to original sources

arXiv2026

Predicting protein structures at any folding stage from sequence

Protein structure predictors reliably identify native conformations, but do not describe the continuum connecting unfolded ensembles to folded states. Here we present FoldScape, a sequence-conditioned deep learning framework that generates C$α$ conformational ensembles at a specified folding fraction $Q$. FoldScape reconstructs three-dimensional chains through hierarchical, sliding-window regressions of one-dimensional inter-residue distances, followed by chain assembly and sampling with a rectified-flow decoder. It is trained on multi-scale coarse-grained folding trajectories and all-atom mechanical-unfolding trajectories. Across 20 human proteins (36-857 residues), at matched sample size, FoldScape covers about an order of magnitude more of the folding coordinate than seed-varied AlphaFold 3 ensembles, robustly across proteins, order parameters and binning choices. AlphaFold 3 better anchors the immediate native vicinity, which FoldScape does not reach, making the approaches complementary. FoldScape reconstruction accuracy remains approximately constant along $Q$, from collapsed coil to near-native globule, rather than deteriorating away from the native state. A stride-one crop-and-superpose scheme extends inference to chains several times longer than a single chain-level forward pass. We also show that training trajectories resolve mechanical unfolding at residue resolution, that elongation-conditioned ensembles exhibit reproducible continuous observables across pulling velocities, and that partially unfolded chains form geometrically coherent segments. FoldScape thus combines broad coverage of folding stages with stable reconstruction accuracy, without per-protein simulation.

q-bio.BM↗

arXiv2026

Foresight: Learning Twisted SMC for Rare Events from Reversed Relaxations

We consider the probability that a reversible Markov chain started in a set $A$ is in a rare set $B$ after a fixed horizon, a quantity often needed for many start laws, down to single states. Twisted sequential Monte Carlo (SMC) estimates such probabilities without bias, with small variance when its proposal and twist approximate the dynamics conditioned on reaching $B$, but these components are usually learned from forward simulations, which contain almost no successful paths when the event is rare. We introduce Foresight-SMC, which learns both components from the target side: under reversibility, relaxations started in $B$ that return to $A$ are, read backwards, exact samples of the conditioned paths, and their number is governed by the return probability rather than by the rare probability. Extending this reversal to the recorded propose--accept decisions of a Metropolis-adjusted sampler, we fit a controlled Langevin proposal with exact likelihood ratios by maximum likelihood, together with a classifier twist; the resulting estimator is unbiased for any start law. A variance analysis shows that resampling reduces the variance when the importance weights degenerate, as they do over long horizons. On an Allen--Cahn field, the learned proposal alone suffices, and one trained model is more accurate than splitting at equal cost. On alanine tripeptide, resampling with the learned twist is necessary, and Foresight-SMC needs less computation per query than adaptive multilevel splitting on localized start sets.

cs.LG↗

arXiv2026

NetCoFi: A Budget-Aware Coarse-Fine Dual-Branch Model for Encrypted Traffic Classification

Encrypted traffic classification is critical for network management and security, but practical deployment must balance classification accuracy and inference cost. Existing models typically process fixed flow windows with limited control over computation after training, which can limit the accuracy-efficiency trade-off under constrained resources. To address this problem, we present NetCoFi, a budget-aware classifier that allocates computation at coarse and fine granularities. Using byte representations and observable packet metadata, NetCoFi selects token neighborhoods for fine-grained encoding while compressing the full window into complementary coarse context. Multi-budget training enables the same model to operate at different routing budgets without retraining. Evaluation on six public datasets shows that NetCoFi achieves state-of-the-art classification performance when trained from scratch. Under low-budget settings, NetCoFi improves average Macro-F1 by 7.86 percentage points across three datasets, compared with a same-depth static baseline. In a DPDK-GPU traffic processing pipeline, it increases mean batch throughput from 26.42 to 43.42 Mb/s, a 1.64x improvement over NetMamba.

cs.NI↗

arXiv2026

A de Sitter Anti-Scrambling Algebra

We study the algebra of observables of a semiclassical observer in de Sitter space. When operators are time-evolved by a scrambling time, out-of-time-ordered correlation functions (OTOCs) receive corrections due to shockwave scattering near the cosmological horizon. When the observer's clock is treated classically, we show that the time advance effect implies the breakdown of the KMS condition for vacuum correlators of matter operators. When the observer's clock is quantized, this implies that the Hartle-Hawking state is not a trace on the resulting crossed-product algebra. We conjecture that the observer's algebra has a trivial commutant. When the time separation between operators exceeds the scrambling time, the OTOC decays to zero and a free product algebra emerges. We illustrate this using classical solutions of de Sitter JT gravity with shocks. We comment on how our results pose a challenge for observer-centric static patch holography.

hep-th↗

arXiv2025

BLADE: Behavior-Level Anomaly Detection Using Network Traffic in Web Services

With their widespread popularity, web services have become the main targets of various cyberattacks. Existing traffic anomaly detection approaches focus on flow-level attacks, yet fail to recognize behavior-level attacks, which appear benign in individual flows but reveal malicious purpose using multiple network flows. To transcend this limitation, we propose a novel unsupervised traffic anomaly detection system, BLADE, capable of detecting not only flow-level but also behavior-level attacks in web services. Our key observation is that application-layer operations of web services exhibit distinctive communication patterns at the network layer from a multi-flow perspective. BLADE first exploits a flow autoencoder to learn a latent feature representation and calculates its reconstruction losses per flow. Then, the latent representation is assigned a pseudo operation label using an unsupervised clustering method. Next, an anomaly score is computed based on the reconstruction losses. Finally, the triplets of timestamps, pseudo labels, and anomaly scores from multiple flows are aggregated and fed into a one-class classifier to characterize the behavior patterns of legitimate web operations, enabling the detection of flow-level and behavior-level anomalies. BLADE is extensively evaluated on both the custom dataset and the CIC-IDS2017 dataset. The experimental results demonstrate BLADE's superior performance, achieving high F1 scores of 0.9732 and 0.9801, respectively, on the two datasets, and outperforming traditional single-flow anomaly detection baselines.

cs.CR↗

arXiv2025

Comprehensive Metapath-based Heterogeneous Graph Transformer for Gene-Disease Association Prediction

Discovering gene-disease associations is crucial for understanding disease mechanisms, yet identifying these associations remains challenging due to the time and cost of biological experiments. Computational methods are increasingly vital for efficient and scalable gene-disease association prediction. Graph-based learning models, which leverage node features and network relationships, are commonly employed for biomolecular predictions. However, existing methods often struggle to effectively integrate node features, heterogeneous structures, and semantic information. To address these challenges, we propose COmprehensive MEtapath-based heterogeneous graph Transformer(COMET) for predicting gene-disease associations. COMET integrates diverse datasets to construct comprehensive heterogeneous networks, initializing node features with BioGPT. We define seven Metapaths and utilize a transformer framework to aggregate Metapath instances, capturing global contexts and long-distance dependencies. Through intra- and inter-metapath aggregation using attention mechanisms, COMET fuses latent vectors from multiple Metapaths to enhance GDA prediction accuracy. Our method demonstrates superior robustness compared to state-of-the-art approaches. Ablation studies and visualizations validate COMET's effectiveness, providing valuable insights for advancing human health research.

cs.AI↗

arXiv2024

Towards Graph Prompt Learning: A Survey and Beyond

Large-scale "pre-train and prompt learning" paradigms have demonstrated remarkable adaptability, enabling broad applications across diverse domains such as question answering, image recognition, and multimodal retrieval. This approach fully leverages the potential of large-scale pre-trained models, reducing downstream data requirements and computational costs while enhancing model applicability across various tasks. Graphs, as versatile data structures that capture relationships between entities, play pivotal roles in fields such as social network analysis, recommender systems, and biological graphs. Despite the success of pre-train and prompt learning paradigms in Natural Language Processing (NLP) and Computer Vision (CV), their application in graph domains remains nascent. In graph-structured data, not only do the node and edge features often have disparate distributions, but the topological structures also differ significantly. This diversity in graph data can lead to incompatible patterns or gaps between pre-training and fine-tuning on downstream graphs. We aim to bridge this gap by summarizing methods for alleviating these disparities. This includes exploring prompt design methodologies, comparing related techniques, assessing application scenarios and datasets, and identifying unresolved problems and challenges. This survey categorizes over 100 relevant works in this field, summarizing general design principles and the latest applications, including text-attributed graphs, molecules, proteins, and recommendation systems. Through this extensive review, we provide a foundational understanding of graph prompt learning, aiming to impact not only the graph mining community but also the broader Artificial General Intelligence (AGI) community.

cs.LG↗

arXiv2023

Recent Progress and Next Steps for the MATHUSLA LLP Detector

We report on recent progress and next steps in the design of the proposed MATHUSLA Long Lived Particle (LLP) detector for the HL-LHC as part of the Snowmass 2021 process. Our understanding of backgrounds has greatly improved, aided by detailed simulation studies, and significant R&D has been performed on designing the scintillator detectors and understanding their performance. The collaboration is on track to complete a Technical Design Report, and there are many opportunities for interested new members to contribute towards the goal of designing and constructing MATHUSLA in time for HL-LHC collisions, which would increase the sensitivity to a large variety of highly motivated LLP signals by orders of magnitude.

hep-ex↗

arXiv2020

An Update to the Letter of Intent for MATHUSLA: Search for Long-Lived Particles at the HL-LHC

We report on recent progress in the design of the proposed MATHUSLA Long Lived Particle (LLP) detector for the HL-LHC, updating the information in the original Letter of Intent (LoI), see CDS:LHCC-I-031, arXiv:1811.00927. A suitable site has been identified at LHC Point 5 that is closer to the CMS Interaction Point (IP) than assumed in the LoI. The decay volume has been increased from 20 m to 25 m in height. Engineering studies have been made in order to locate much of the decay volume below ground, bringing the detector even closer to the IP. With these changes, a 100 m x 100 m detector has the same physics reach for large c$τ$ as the 200 m x 200 m detector described in the LoI and other studies. The performance for small c$τ$ is improved because of the proximity to the IP. Detector technology has also evolved while retaining the strip-like sensor geometry in Resistive Plate Chambers (RPC) described in the LoI. The present design uses extruded scintillator bars read out using wavelength shifting fibers and silicon photomultipliers (SiPM). Operations will be simpler and more robust with much lower operating voltages and without the use of greenhouse gases. Manufacturing is straightforward and should result in cost savings. Understanding of backgrounds has also significantly advanced, thanks to new simulation studies and measurements taken at the MATHUSLA test stand operating above ATLAS in 2018. We discuss next steps for the MATHUSLA collaboration, and identify areas where new members can make particularly important contributions.

physics.ins-det↗