arXiv subjects

Yiping Liu

Publications and source records attributed to Yiping Liu.

At least 19 recordsLinked to original sources

arXiv2026

CryoProt: A Protein Pretraining Framework with Cross-Box Interactions on Cryo-EM Density Maps

Despite the growing availability of cryo-electron microscopy (cryo-EM) density maps, effectively leveraging them for protein representation remains challenging. First, current methods lack a general-purpose protein pretraining framework tailored for cryo-EM density maps, designed for protein-related property prediction. Second, existing approaches typically partition density maps into local box regions and model them independently, overlooking interactions across boxes which are essential for capturing global structural context in cryo-EM density map. To address these challenges, we propose CryoProt, a protein pretraining framework designed for cryo-EM density maps. CryoProt introduces a Map Encoder based on multi-head latent attention (MLA), where box-level representations interact through a shared latent space, enabling explicit modeling of cross-box dependencies within the density map. Furthermore, we adopt a multi-task pretraining strategy to learn generalizable representations that can be effectively transferred to diverse downstream tasks, such as protein flexibility prediction, where cryo-EM density maps are not required and can be inferred implicitly by the pretrained model. Experimental results demonstrate that CryoProt consistently outperforms existing state-of-the-art methods across multiple benchmarks, achieving up to 12% improvement over the best-performing baselines, highlighting the importance of modeling cross-box interactions in cryo-EM data. The source code is publicly available at https://anonymous.4open.science/r/CryoProt.

cs.LG↗

arXiv2026

Data-Driven Privacy-Preserving Modeling and Frequency Regulation with Aggregated Electric Vehicles via Bilinear Hidden Markov Model

Vehicle-to-Grid (V2G) technology allows bidirectional power flow for real-time grid support, making electric vehicles (EVs) well-suited for ancillary services such as frequency regulation. However, existing methods for flexibility estimation and coordinating aggregated EVs often rely on individual EV traveling information (e.g., arrival/departure time) and/or characteristic parameters (e.g., charging efficiency, battery capacity) as well as real-time state-of-charge (SOC), which raises privacy concerns and faces data quality issues. To address these challenges, this paper proposes a data-driven, privacy-preserving modeling and control framework for frequency regulation using aggregated EVs. The proposed method can provide accurate estimation for power outputs and flexibility of aggregated EVs and carry out effective frequency regulation without any individual EV information. Simulation results validate the accuracy and effectiveness of the proposed method, which also outperforms the model-based and federated learning-based method under SOC data inaccuracies.

eess.SY↗

arXiv2026

Generalized parton distributions of valence, sea, and gluon components of the proton

We compute the generalized parton distributions (GPDs) of valence quarks, sea quarks, and gluons in the proton using light-front wave functions obtained within the basis light-front quantization (BLFQ) framework, providing a realistic description of the nucleon at a low resolution scale. The wave functions are derived from a light-front QCD Hamiltonian without an explicit confining potential and include the three-quark, three-quark-gluon, and three-quark-quark-antiquark Fock sectors. For the first time within BLFQ, we evaluate quark GPDs at nonzero skewness in both the DGLAP and ERBL regions, while gluon GPDs are computed in the DGLAP region. The resulting GPDs exhibit qualitative features similar to, but smaller than the GUMP1.0 global extraction of GPDs based on experimental and lattice QCD data at next-to-leading order accuracy. We further compute the associated Compton form factors and obtain results consistent with the global analysis.

hep-ph↗

arXiv2026

Sparse Task Vector Mixup with Hypernetworks for Efficient Knowledge Transfer in Whole-Slide Image Prognosis

Whole-Slide Images (WSIs) are widely used for estimating the prognosis of cancer patients. Current studies generally follow a cancer-specific learning paradigm. However, the available training samples for one cancer type are usually scarce in pathology. Consequently, the model often struggles to learn generalizable knowledge, thus performing worse on the tumor samples with inherent high heterogeneity. Although multi-cancer joint learning and knowledge transfer approaches have been explored recently to address it, they either rely on large-scale joint training or extensive inference across multiple models, posing new challenges in computational efficiency. To this end, this paper proposes a new scheme, Sparse Task Vector Mixup with Hypernetworks (STEPH). Unlike previous ones, it efficiently absorbs generalizable knowledge from other cancers for the target via model merging: i) applying task vector mixup to each source-target pair and then ii) sparsely aggregating task vector mixtures to obtain an improved target model, driven by hypernetworks. Extensive experiments on 13 cancer datasets show that STEPH improves over cancer-specific learning and an existing knowledge transfer baseline by 5.14% and 2.01%, respectively. Moreover, it is a more efficient solution for learning prognostic knowledge from other cancers, without requiring large-scale joint training or extensive multi-model inference. Code is publicly available at https://github.com/liupei101/STEPH.

cs.CV↗

arXiv2026

DyACE: Dynamic Algorithm Co-evolution for Online Automated Heuristic Design with Large Language Model

The prevailing paradigm in Automated Heuristic Design (AHD) typically relies on the assumption that a single, fixed algorithm can effectively navigate the shifting dynamics of a combinatorial search. This static approach often proves inadequate for Perturbative Heuristics, where the optimal algorithm for escaping local optima depends heavily on the specific search phase. To address this limitation, we reformulate heuristic design as a Non-stationary Bi-level Control problem and introduce DyACE (Dynamic Algorithm Co-evolution). Distinct from standard open-loop solvers, DyACE use a Receding Horizon Control architecture to continuously co-evolve the heuristic logic alongside the solution population. A core element of this framework is the Look-Ahead Rollout Search, which queries the landscape geometry to extract Search Trajectory Features. This sensory feedback allows the Large Language Model (LLM) to function as a grounded meta-controller, prescribing phase-specific interventions tailored to the real-time search status. We validate DyACE on three representative combinatorial optimization benchmarks. The results demonstrate that our method significantly outperforms state-of-the-art static baselines, exhibiting superior scalability in high-dimensional search spaces. Furthermore, ablation studies confirm that dynamic adaptation fails without grounded perception, often performing worse than static algorithms. This indicates that DyACE's effectiveness stems from the causal alignment between the synthesized logic and the verified gradients of the optimization landscape.

cs.AI↗

arXiv2026

MergeDJD: A Fast Constructive Algorithm with Piece Merging for the Two-Dimensional Irregular Bin Packing Problem

The two-dimensional irregular bin packing problem (2DIBPP) aims to pack a given set of irregular polygons, referred to as pieces, into fixed-size rectangular bins without overlap, while maximizing bin utilization. Although numerous metaheuristic algorithms have been proposed for the 2DIBPP, many industrial applications favor simpler constructive heuristics due to their deterministic behavior and low computational overhead. Among such methods, the DJD algorithm proposed by L'opez-Camacho et al. is one of the most competitive constructive heuristics for the 2DIBPP. However, DJD is less effective for cutting instances, in which many pieces can be seamlessly combined into larger polygons. To address the issue, we propose MergeDJD, a novel constructive algorithm that integrates and extends the DJD framework. MergeDJD first preprocesses the instance by iteratively identifying groups of pieces that can be combined into larger and more regular piece. It then employs an improved version of DJD, in which the placement strategy is enhanced to better handle non-convex and combined shapes, to pack all resulting pieces into bins. Computational experiments on 1,089 well-known benchmark instances show that MergeDJD consistently outperforms DJD on 1,083 instances while maintaining short runtimes. Notably, MergeDJD attains new best known values on 515 instances. Ablation studies further confirm the effectiveness of the proposed components. To facilitate reproducibility and future research, we have open-sourced the complete implementation and provided interfaces for visualizing packing results.

cs.CG↗

arXiv2025

Retrofit: Continual Learning with Controlled Forgetting for Binary Security Detection and Analysis

Binary security has increasingly relied on deep learning to reason about malware behavior and program semantics. However, the performance often degrades as threat landscapes evolve and code representations shift. While continual learning (CL) offers a natural solution through sequential updates, most existing approaches rely on data replay or unconstrained updates, limiting their applicability and effectiveness in data-sensitive security environments. We propose RETROFIT, which regulates knowledge retention and adaptation with controlled forgetting at each update, without requiring historical data. Our key idea is to consolidate previously trained and newly fine-tuned models, serving as teachers of legacy and emergent knowledge, through retrospective-free parameter merging. Forgetting control is achieved by 1) constraining parameter changes to low-rank and sparse subspaces for approximate orthogonality, and 2) employing a confidence-guided arbitration mechanism to dynamically aggregate knowledge from both teachers. Our evaluation on two representative applications demonstrates that RETROFIT consistently mitigates forgetting while maintaining adaptability. In malware detection under temporal drift, it substantially improves the retention score, from 20.2% to 38.6% over CL baselines, and exceeds the oracle upper bound on new data. In binary summarization across decompilation levels, where analyzing stripped binaries is especially challenging, RETROFIT achieves over 2x the BLEU score of transfer learning used in prior work and surpasses all baselines in cross-representation generalization.

cs.LG↗

arXiv2025

DynaQuant: Dynamic Mixed-Precision Quantization for Learned Image Compression

Prevailing quantization techniques in Learned Image Compression (LIC) typically employ a static, uniform bit-width across all layers, failing to adapt to the highly diverse data distributions and sensitivity characteristics inherent in LIC models. This leads to a suboptimal trade-off between performance and efficiency. In this paper, we introduce DynaQuant, a novel framework for dynamic mixed-precision quantization that operates on two complementary levels. First, we propose content-aware quantization, where learnable scaling and offset parameters dynamically adapt to the statistical variations of latent features. This fine-grained adaptation is trained end-to-end using a novel Distance-aware Gradient Modulator (DGM), which provides a more informative learning signal than the standard Straight-Through Estimator. Second, we introduce a data-driven, dynamic bit-width selector that learns to assign an optimal bit precision to each layer, dynamically reconfiguring the network's precision profile based on the input data. Our fully dynamic approach offers substantial flexibility in balancing rate-distortion (R-D) performance and computational cost. Experiments demonstrate that DynaQuant achieves rd performance comparable to full-precision models while significantly reducing computational and storage requirements, thereby enabling the practical deployment of advanced LIC on diverse hardware platforms.

eess.IV↗

arXiv2025

From Knowledge to Treatment: Large Language Model Assisted Biomedical Concept Representation for Drug Repurposing

Drug repurposing plays a critical role in accelerating treatment discovery, especially for complex and rare diseases. Biomedical knowledge graphs (KGs), which encode rich clinical associations, have been widely adopted to support this task. However, existing methods largely overlook common-sense biomedical concept knowledge in real-world labs, such as mechanistic priors indicating that certain drugs are fundamentally incompatible with specific treatments. To address this gap, we propose LLaDR, a Large Language Model-assisted framework for Drug Repurposing, which improves the representation of biomedical concepts within KGs. Specifically, we extract semantically enriched treatment-related textual representations of biomedical entities from large language models (LLMs) and use them to fine-tune knowledge graph embedding (KGE) models. By injecting treatment-relevant knowledge into KGE, LLaDR largely improves the representation of biomedical concepts, enhancing semantic understanding of under-studied or complex indications. Experiments based on benchmarks demonstrate that LLaDR achieves state-of-the-art performance across different scenarios, with case studies on Alzheimer's disease further confirming its robustness and effectiveness. Code is available at https://github.com/xiaomingaaa/LLaDR.

cs.CL↗

arXiv2025

ImageDDI: Image-enhanced Molecular Motif Sequence Representation for Drug-Drug Interaction Prediction

To mitigate the potential adverse health effects of simultaneous multi-drug use, including unexpected side effects and interactions, accurately identifying and predicting drug-drug interactions (DDIs) is considered a crucial task in the field of deep learning. Although existing methods have demonstrated promising performance, they suffer from the bottleneck of limited functional motif-based representation learning, as DDIs are fundamentally caused by motif interactions rather than the overall drug structures. In this paper, we propose an Image-enhanced molecular motif sequence representation framework for \textbf{DDI} prediction, called ImageDDI, which represents a pair of drugs from both global and local structures. Specifically, ImageDDI tokenizes molecules into functional motifs. To effectively represent a drug pair, their motifs are combined into a single sequence and embedded using a transformer-based encoder, starting from the local structure representation. By leveraging the associations between drug pairs, ImageDDI further enhances the spatial representation of molecules using global molecular image information (e.g. texture, shadow, color, and planar spatial relationships). To integrate molecular visual information into functional motif sequence, ImageDDI employs Adaptive Feature Fusion, enhancing the generalization of ImageDDI by dynamically adapting the fusion process of feature representations. Experimental results on widely used datasets demonstrate that ImageDDI outperforms state-of-the-art methods. Moreover, extensive experiments show that ImageDDI achieved competitive performance in both 2D and 3D image-enhanced scenarios compared to other models.

cs.CV↗

arXiv2025

Kimi K2: Open Agentic Intelligence

We introduce Kimi K2, a Mixture-of-Experts (MoE) large language model with 32 billion activated parameters and 1 trillion total parameters. We propose the MuonClip optimizer, which improves upon Muon with a novel QK-clip technique to address training instability while enjoying the advanced token efficiency of Muon. Based on MuonClip, K2 was pre-trained on 15.5 trillion tokens with zero loss spike. During post-training, K2 undergoes a multi-stage post-training process, highlighted by a large-scale agentic data synthesis pipeline and a joint reinforcement learning (RL) stage, where the model improves its capabilities through interactions with real and synthetic environments. Kimi K2 achieves state-of-the-art performance among open-source non-thinking models, with strengths in agentic capabilities. Notably, K2 obtains 66.1 on Tau2-Bench, 76.5 on ACEBench (En), 65.8 on SWE-Bench Verified, and 47.3 on SWE-Bench Multilingual -- surpassing most open and closed-sourced baselines in non-thinking settings. It also exhibits strong capabilities in coding, mathematics, and reasoning tasks, with a score of 53.7 on LiveCodeBench v6, 49.5 on AIME 2025, 75.1 on GPQA-Diamond, and 27.1 on OJBench, all without extended thinking. These results position Kimi K2 as one of the most capable open-source large language models to date, particularly in software engineering and agentic tasks. We release our base and post-trained model checkpoints to facilitate future research and applications of agentic intelligence.

cs.LG↗

arXiv2025

Dataset Distillation as Data Compression: A Rate-Utility Perspective

Driven by the ``scale-is-everything'' paradigm, modern machine learning increasingly demands ever-larger datasets and models, yielding prohibitive computational and storage requirements. Dataset distillation mitigates this by compressing an original dataset into a small set of synthetic samples, while preserving its full utility. Yet, existing methods either maximize performance under fixed storage budgets or pursue suitable synthetic data representations for redundancy removal, without jointly optimizing both objectives. In this work, we propose a joint rate-utility optimization method for dataset distillation. We parameterize synthetic samples as optimizable latent codes decoded by extremely lightweight networks. We estimate the Shannon entropy of quantized latents as the rate measure and plug any existing distillation loss as the utility measure, trading them off via a Lagrange multiplier. To enable fair, cross-method comparisons, we introduce bits per class (bpc), a precise storage metric that accounts for sample, label, and decoder parameter costs. On CIFAR-10, CIFAR-100, and ImageNet-128, our method achieves up to $170\times$ greater compression than standard distillation at comparable accuracy. Across diverse bpc budgets, distillation losses, and backbone architectures, our approach consistently establishes better rate-utility trade-offs.

cs.LG↗

arXiv2025

Numerical Linear Algebra in Linear Space

We present a randomized linear-space solver for general linear systems $\mathbf{A} \mathbf{x} = \mathbf{b}$ with $\mathbf{A} \in \mathbb{Z}^{n \times n}$ and $\mathbf{b} \in \mathbb{Z}^n$, without any assumption on the condition number of $\mathbf{A}$. For matrices whose entries are bounded by $\mathrm{poly}(n)$, the solver returns a $(1+\epsilon)$-multiplicative entry-wise approximation to vector $\mathbf{x} \in \mathbb{Q}^{n}$ using $\widetilde{O}(n^2 \cdot \mathrm{nnz}(\mathbf{A}))$ bit operations and $O(n \log n)$ bits of working space (i.e., linear in the size of a vector), where $\mathrm{nnz}$ denotes the number of nonzero entries. Our solver works for right-hand vector $\mathbf{b}$ with entries up to $n^{O(n)}$. To our knowledge, this is the first linear-space linear system solver over the rationals that runs in $\widetilde{O}(n^2 \cdot \mathrm{nnz}(\mathbf{A}))$ time. We also present several applications of our solver to numerical linear algebra problems, for which we provide algorithms with efficient polynomial running time and near-linear space. In particular, we present results for linear regression, linear programming, eigenvalues and eigenvectors, and singular value decomposition.

cs.DS↗

arXiv2025

OJBench: A Competition Level Code Benchmark For Large Language Models

Recent advancements in large language models (LLMs) have demonstrated significant progress in math and code reasoning capabilities. However, existing code benchmark are limited in their ability to evaluate the full spectrum of these capabilities, particularly at the competitive level. To bridge this gap, we introduce OJBench, a novel and challenging benchmark designed to assess the competitive-level code reasoning abilities of LLMs. OJBench comprises 232 programming competition problems from NOI and ICPC, providing a more rigorous test of models' reasoning skills. We conducted a comprehensive evaluation using OJBench on 37 models, including both closed-source and open-source models, reasoning-oriented and non-reasoning-oriented models. Our results indicate that even state-of-the-art reasoning-oriented models, such as o4-mini and Gemini-2.5-pro-exp, struggle with highly challenging competition-level problems. This highlights the significant challenges that models face in competitive-level code reasoning.

cs.CL↗

arXiv2025

A Reduction-Driven Local Search for the Generalized Independent Set Problem

The Generalized Independent Set (GIS) problem extends the classical maximum independent set problem by incorporating profits for vertices and penalties for edges. This generalized problem has been identified in diverse applications in fields such as forest harvest planning, competitive facility location, social network analysis, and even machine learning. However, solving the GIS problem in large-scale, real-world networks remains computationally challenging. In this paper, we explore data reduction techniques to address this challenge. We first propose 14 reduction rules that can reduce the input graph with rigorous optimality guarantees. We then present a reduction-driven local search (RLS) algorithm that integrates these reduction rules into the pre-processing, the initial solution generation, and the local search components in a computationally efficient way. The RLS is empirically evaluated on 278 graphs arising from different application scenarios. The results indicates that the RLS is highly competitive -- For most graphs, it achieves significantly superior solutions compared to other known solvers, and it effectively provides solutions for graphs exceeding 260 million edges, a task at which every other known method fails. Analysis also reveals that the data reduction plays a key role in achieving such a competitive performance.

cs.IR↗

arXiv2025

AdaptMol: Adaptive Fusion from Sequence String to Topological Structure for Few-shot Drug Discovery

Accurate molecular property prediction (MPP) is a critical step in modern drug development. However, the scarcity of experimental validation data poses a significant challenge to AI-driven research paradigms. Under few-shot learning scenarios, the quality of molecular representations directly dictates the theoretical upper limit of model performance. We present AdaptMol, a prototypical network integrating Adaptive multimodal fusion for Molecular representation. This framework employs a dual-level attention mechanism to dynamically integrate global and local molecular features derived from two modalities: SMILES sequences and molecular graphs. (1) At the local level, structural features such as atomic interactions and substructures are extracted from molecular graphs, emphasizing fine-grained topological information; (2) At the global level, the SMILES sequence provides a holistic representation of the molecule. To validate the necessity of multimodal adaptive fusion, we propose an interpretable approach based on identifying molecular active substructures to demonstrate that multimodal adaptive fusion can efficiently represent molecules. Extensive experiments on three commonly used benchmarks under 5-shot and 10-shot settings demonstrate that AdaptMol achieves state-of-the-art performance in most cases. The rationale-extracted method guides the fusion of two modalities and highlights the importance of both modalities.

cs.LG↗

arXiv2025

Basis light-front quantization: Advancing a first principles approach for the nucleon

We present our recent progress in applying the basis light-front quantization approach to investigate the nucleon's structure. We solve its wave functions from the eigenstates of the light-front QCD Hamiltonian using a fully relativistic, nonperturbative approach without an explicit confining potential. These eigenstates are determined for the three-quark, three-quark-gluon, and three-quark-quark-antiquark Fock representations, making them suitable for low-resolution probes. From these wave functions, we compute the nucleon's quark and gluon matter densities, as well as its helicity and transversity distributions, showing qualitative consistency with experimental data. We also determine the quark and gluon helicity contributions to the proton spin and the tensor charges. The resulting light-front wave functions represent a significant step toward a unified description of hadron distribution functions in both longitudinal and transverse momentum space.

hep-ph↗

arXiv2025

An Extended State Space Model of Aggregated Electric Vehicles for Flexibility Estimation and Power Control

The increasing penetration of electric vehicles (EVs) can provide substantial electricity to the grid, supporting the grids' stability. The state space model (SSM) has been proposed as an effective modeling method for power prediction and centralized control of aggregated EVs, offering low communication requirements and computational complexity. However, the SSM may overlook specific scenarios, leading to significant prediction and control inaccuracies. This paper proposes an extended state space model (eSSM) for aggregated EVs and develops associated control strategies. By accounting for the limited flexibility of fully charged and discharged EVs, the eSSM more accurately captures the state transition dynamics of EVs in various states of charge (SOC). Comprehensive simulations show that the eSSM will provide more accurate predictions of the flexibility and power trajectories of aggregated EVs, and more effectively tracks real-time power references compared to the conventional SSM method.

eess.SY↗