Search arXiv⌕ Search

arXiv subjects

Yang Wang

Publications and source records attributed to Yang Wang.

At least 145 records · Page 8Linked to original sources
arXiv2026

PepBenchmark: A Standardized Benchmark for Peptide Machine Learning

Peptide therapeutics are widely regarded as the "third generation" of drugs, yet progress in peptide Machine Learning (ML) are hindered by the absence of standardized benchmarks. Here we present PepBenchmark, which unifies datasets, preprocessing, and evaluation protocols for peptide drug discovery. PepBenchmark comprises three components: (1) PepBenchData, a well-curated collection comprising 29 canonical-peptide and 6 non-canonical-peptide datasets across 7 groups, systematically covering key aspects of peptide drug development, representing, to the best of our knowledge, the most comprehensive AI-ready dataset resource to date; (2) PepBenchPipeline, a standardized preprocessing pipeline that ensures consistent dataset cleaning, construction, splitting, and feature transformation, mitigating quality issues common in ad hoc pipelines; and (3) PepBenchLeaderboard, a unified evaluation protocol and leaderboard with strong baselines across 4 major methodological families: Fingerprint-based, GNN-based, PLM-based, and SMILES-based models. Together, PepBenchmark provides the first standardized and comparable foundation for peptide drug discovery, facilitating methodological advances and translation into real-world applications. The data and code are publicly available at https://github.com/ZGCI-AI4S-Pep/PepBenchmark/.

cs.LG↗
arXiv2026

Enhanced Self-Learning with Epistemologically-Informed LLM Dialogue

Large Language Models (LLMs) have advanced self-learning tools, enabling more personalized interactions. However, learners struggle to engage in meaningful dialogue and process complex information. To alleviate this, we incorporate epistemological frameworks within an LLM-based approach to self-learning, reducing the cognitive load on learners and fostering deeper engagement and holistic understanding. Through a formative study (N=26), we identified epistemological differences in self-learner interaction patterns. Building upon these findings, we present \textit{CausaDisco}, a dialogue-based interactive system that integrates Aristotle's \textit{Four Causes} framework into LLM prompts to enhance cognitive support for self-learning. This approach guides learners' self-learning journeys by automatically generating coherent and contextually appropriate follow-up questions. A controlled study (N=36) demonstrated that, compared to baseline, \textit{CausaDisco} fostered more engaging interactions, inspired sophisticated exploration, and facilitated multifaceted perspectives. This research contributes to HCI by expanding the understanding of LLMs as educational agents and providing design implications for this emerging class of tools.

cs.HC↗
arXiv2026

Deep Photonic Reservoir Computer Meets UAV Control: An ultra-fast learning-based compensator for agile flight in confined space

Unmanned aerial vehicles (UAVs) operating in confined, cluttered environments face significant performance degradation due to nonlinear, time-varying unmodeled dynamics-such as ground/ceiling effects and wake recirculation-that are unaccounted for in traditional controllers. While learning based compensators (e.g., MLPs, TCNs, LSTMs) struggle with historical data dependency, vanishing gradients, and prohibitive computational costs, this work pioneers the integration of a deep photonic reservoir computer (PRC) with feedforward control to overcome these limitations. Harnessing semiconductor laser dynamics and optical feedback, our hardware implemented deep PRC architecture achieves intrinsic temporal memory without explicit historical inputs, while reducing training time from hours to milliseconds and slashing inference latency to nanoseconds. Reliable high-performance CFD simulations capturing proximity-induced flows demonstrate that deep PRC delivers residual-force prediction accuracy comparable to or exceeding TCN/MLP baselines, while training only a linear readout layer via ridge regression. By injecting these predictions into a nonlinear feedback PID controller via a feedforward channel, the framework significantly enhances closed-loop tracking stability in confined spaces. Essentially, this work establishes the first deep PRC-based lightweight, ultrafast solution for real-time UAV dynamic compensation, with promising extensibility to unseen scenarios with more complex fluid environments.

physics.optics↗
arXiv2026

VAG: Dual-Stream Video-Action Generation for Embodied Data Synthesis

Recent advances in robot foundation models trained on large-scale human teleoperation data have enabled robots to perform increasingly complex real-world tasks. However, scaling these systems remains difficult because collecting task-specific demonstrations is expensive and labor-intensive. Synthetic data, especially generated videos, offer a promising direction, but existing World Models (WMs) are not directly suitable for policy learning since they do not provide paired action trajectories. World-Action (WA) models partially address this by predicting actions with visual outputs, yet often lack strong video-action alignment, while two-stage pipelines that generate video first and then infer actions introduce inefficiency and error accumulation. To address these limitations, we propose VAG, a unified flow-matching-based dual-stream framework that jointly generates video and action under visual and language conditioning. By synchronizing denoising in both branches and using an adaptive 3D pooling mechanism to transfer compact global video context to the action branch, VAG improves cross-modal consistency during generation. Across both simulated and real-world settings, VAG produces aligned video-action pairs with competitive prediction quality, supports executable trajectory replay, and provides useful synthetic pretraining data that improves downstream policy generalization, indicating its potential as a practical world-action model for embodied data synthesis.

cs.RO↗
arXiv2026

ETR: Entropy Trend Reward for Efficient Chain-of-Thought Reasoning

Chain-of-thought (CoT) reasoning improves large language model performance on complex tasks, but often produces excessively long and inefficient reasoning traces. Existing methods shorten CoTs using length penalties or global entropy reduction, implicitly assuming that low uncertainty is desirable throughout reasoning. We show instead that reasoning efficiency is governed by the trajectory of uncertainty. CoTs with dominant downward entropy trends are substantially shorter. Motivated by this insight, we propose Entropy Trend Reward (ETR), a trajectory-aware objective that encourages progressive uncertainty reduction while allowing limited local exploration. We integrate ETR into Group Relative Policy Optimization (GRPO) and evaluate it across multiple reasoning models and challenging benchmarks. ETR consistently achieves a superior accuracy-efficiency tradeoff, improving DeepSeek-R1-Distill-7B by 9.9% in accuracy while reducing CoT length by 67% across four benchmarks. Code is available at https://github.com/Xuan1030/ETR

cs.AI↗
arXiv2026

Non-Markovian Entropy Dynamics in Living Systems from the Keldysh Formalism

Living systems are open nonequilibrium systems that continuously exchange energy, matter, and information with their environments, leading to stochastic dynamics with memory and active fluctuations. In this study, we develop a non-Markovian theoretical framework for the entropy dynamics of living systems based on the Keldysh functional formalism and stochastic thermodynamics. The approach naturally incorporates colored environmental noise, memory-dependent dissipation, and many-body interactions, yielding generalized Langevin dynamics and non-Markovian master equations. Within this framework we derive an exact frequency-domain expression for the entropy production rate and show that violations of the fluctuation-dissipation relation provide a direct thermodynamic signature of active biological fluctuations. We further demonstrate that environmental memory enhances low-frequency fluctuations and entropy production, leading to critical slowing down near dynamical instability. These results provide a microscopic physical foundation for the entropy "bathtub" picture of living systems and connect entropy evolution with development, aging, and death in nonequilibrium dynamics.

cond-mat.stat-mech↗
arXiv2026

TSPO: Breaking the Double Homogenization Dilemma in Multi-turn Search Policy Optimization

Multi-turn tool-integrated reasoning enables Large Language Models (LLMs) to solve complex tasks through iterative information retrieval. However, current reinforcement learning (RL) frameworks for search-augmented reasoning predominantly rely on sparse outcome-level rewards, leading to a "Double Homogenization Dilemma." This manifests as (1) Process homogenization, where the thinking, reasoning, and tooling involved in generation are ignored. (2) Intra-group homogenization, coarse-grained outcome rewards often lead to inefficiencies in intra-group advantage estimation with methods like Group Relative Policy Optimization (GRPO) during sampling. To address this, we propose Turn-level Stage-aware Policy Optimization (TSPO). TSPO introduces the First-Occurrence Latent Reward (FOLR) mechanism, allocating partial rewards to the step where the ground-truth answer first appears, thereby preserving process-level signals and increasing reward variance within groups without requiring external reward models or any annotations. Extensive experiments demonstrate that TSPO significantly outperforms state-of-the-art baselines, achieving average performance gains of 24% and 13.6% on Qwen2.5-3B and 7B models, respectively. Code is available at https://github.com/Flipped-May/TSPO.

cs.AI↗
arXiv2026

Cognibit: From Digital Exhaustion to Real-World Connection Through Gamified Territory Control and LLM-Powered Twin Networking

We present an LLM-powered social discovery platform that uses digital twins to autonomously evaluate interpersonal compatibility through behavioral simulation. The platform unifies three key pillars: (1) digital twins that engage in autonomous multi-turn conversations on behalf of users to estimate compatibility, (2) gamified territory conquest mechanics that incentivize real-world exploration and create organic settings for in-person encounters, and (3) AI companions that preserve persistent shared memory across devices. Built upon CogniPair's cognitive architecture (Ye et al., 2026), validated on the Columbia Speed Dating dataset (551 participants), our system extends prior simulation-only matching into a fully deployed social discovery environment. Through deployment, we derive empirical cost-quality baselines and identify fundamental scaling bottlenecks that remain hidden in component-level testing alone.

cs.HC↗
arXiv2026

RDFace: A Benchmark Dataset for Rare Disease Facial Image Analysis under Extreme Data Scarcity and Phenotype-Aware Synthetic Generation

Rare diseases often manifest with distinctive facial phenotypes in children, offering valuable diagnostic cues for clinicians and AI-assisted screening systems. However, progress in this field is severely limited by the scarcity of curated, ethically sourced facial data and the high similarity among phenotypes across different conditions. To address these challenges, we introduce RDFace, a curated benchmark dataset comprising 456 pediatric facial images spanning 103 rare genetic conditions (average 4.4 samples per condition). Each ethically verified image is paired with standardized metadata. RDFace enables the development and evaluation of data-efficient AI models for rare disease diagnosis under real-world low-data constraints. We benchmark multiple pretrained vision backbones using cross-validation and explore synthetic augmentation with DreamBooth and FastGAN. Generated images are filtered via facial landmark similarity to maintain phenotype fidelity and merged with real data, improving diagnostic accuracy by up to 13.7% in ultra-low-data regimes. To assess semantic validity, phenotype descriptions generated by a vision-language model from real and synthetic images achieve a report similarity score of 0.84. RDFace establishes a transparent, benchmark-ready dataset for equitable rare disease AI research and presents a scalable framework for evaluating both diagnostic performance and the integrity of synthetic medical imagery.

cs.CV↗
arXiv2026

FACT-GS: Frequency-Aligned Complexity-Aware Texture Reparameterization for 2D Gaussian Splatting

Realistic scene appearance modeling has advanced rapidly with Gaussian Splatting, which enables real-time, high-quality rendering. Recent advances introduced per-primitive textures that incorporate spatial color variations within each Gaussian, improving their expressiveness. However, texture-based Gaussians parameterize appearance with a uniform per-Gaussian sampling grid, allocating equal sampling density regardless of local visual complexity, which leads to inefficient texture space utilization. We introduce FACT-GS, a Frequency-Aligned Complexity-aware Texture Gaussian Splatting framework that allocates texture sampling density according to local visual frequency. Grounded in adaptive sampling theory, FACT-GS reformulates texture parameterization as a differentiable sampling-density allocation problem, replacing the uniform textures with a learnable frequency-aware allocation strategy implemented via a deformation field whose Jacobian modulates local sampling density. Built on 2D Gaussian Splatting, FACT-GS performs non-uniform sampling on fixed-resolution texture grids, preserving real-time performance while recovering sharper high-frequency details under the same parameter budget.

cs.CV↗
arXiv2026

Auto-Slides: An Interactive Multi-Agent System for Creating and Customizing Research Presentations

The rapid progress of large language models (LLMs) has opened new opportunities for education. While learners can interact with academic papers through LLM-powered dialogue, limitations still exist: the lack of structured organization and the heavy reliance on text can impede systematic understanding and engagement with complex concepts. To address these challenges, we propose Auto-Slides, an LLM-driven system that converts research papers into pedagogically structured, multimodal slides (e.g., diagrams and tables). Drawing on cognitive science, it creates a presentation-oriented narrative and allows iterative refinement via an interactive editor to better match learners' knowledge level and goals. Auto-Slides further incorporates verification and knowledge retrieval mechanisms to ensure accuracy and contextual completeness. Through extensive user studies, Auto-Slides demonstrates strong learner acceptance, improved structural support for understanding, and expert-validated gains in narrative quality compared with conventional LLM-based reading. Our contributions lie in designing a multi-agent framework for transforming academic papers into pedagogically optimized slides and introducing interactive customization for personalized learning.

cs.HC↗
arXiv2026

A Precision Emulation Approach to the GPU Acceleration of Ab Initio Electronic Structure Calculations

This study explores the use of INT8-based emulation for accelerating traditional FP64-based HPC workloads on modern GPU architectures. Through SCILIB-Accel automatic BLAS offload tool for cache-coherent Unified Memory Architecture, we emulate FP64 matrix multiplications in the LSMS CPU application in the MuST suite without code changes. We find that accuracy depends on both arithmetic precision and the properties of the operator, which can be dealt with through tunable precision emulation. Unlike traditional mixed-precision approaches, this method preserves original algorithms while optimizing hardware utilization. We showcase the potential of improving accuracy and performance at the same time. This work highlights the potential of AI-driven hardware to transform HPC, advocating for adaptive precision strategies in future scientific computing.

cs.DC↗
arXiv2026

LH2Face: Loss function for Hard High-quality Face

In current practical face authentication systems, most face recognition (FR) algorithms are based on cosine similarity with softmax classification. Despite its reliable classification performance, this method struggles with hard samples. A popular strategy to improve FR performance is incorporating angular or cosine margins. However, it does not take face quality or recognition hardness into account, simply increasing the margin value and thus causing an overly uniform training strategy. To address this problem, a novel loss function is proposed, named Loss function for Hard High-quality Face (LH2Face). Firstly, a similarity measure based on the von Mises-Fisher (vMF) distribution is stated, specifically focusing on the logarithm of the Probability Density Function (PDF), which represents the distance between a probability distribution and a vector. Then, an adaptive margin-based multi-classification method using softmax, called the Uncertainty-Aware Margin Function, is implemented in the article. Furthermore, proxy-based loss functions are used to apply extra constraints between the proxy and sample to optimize their representation space distribution. Finally, a renderer is constructed that optimizes FR through face reconstruction and vice versa. Our LH2Face is superior to similiar schemes on hard high-quality face datasets, achieving 49.39% accuracy on the IJB-B dataset, which surpasses the second-place method by 2.37%.

cs.CV↗
arXiv2026

Decoherence-Free Qubit and Chiral Emission from a Giant Molecule in Waveguide QED

Combining decoherence protection with directional photon emission in a single waveguide quantum electrodynamics (QED) device remains an open challenge. Here we show that an artificial giant molecule -- strongly interacting artificial atoms coupled to a photonic waveguide at multiple spatially separated points -- achieves both: a fully operational decoherence-free (DF) qubit and state-dependent chiral single-photon emission, arising from the same photon-interference mechanism. Initialization reduces to a local excitation of a single atom, universal single-qubit gates are implemented by modulating a single atomic frequency, and readout exploits state-dependent chiral emission with directionality reaching 100% and low measurement error of 1.2%. The coexistence of decoherence protection and directional emission in a single device positions giant molecules as protected chiral nodes for modular quantum networks in waveguide QED.

quant-ph↗
arXiv2026

Quadratic shift-and-stack for Ground-Based Optical Detection of Faint Cislunar Objects

Detecting faint objects in cislunar space using ground-based optical telescopes is difficult because of their low brightness, strong lunar background, and complex, nonlinear apparent motion. Traditional shift-and-stack techniques based on linear motion assumption suffer signal trailing loss due to significant nonlinear motion during long integrations, thus producing a degraded signal-to-noise ratio (SNR). In this paper, we first derive a theoretical criterion based on the point spread function to determine the maximum applicable integration time for linear-motion stacking. We then propose a quadratic shift-and-stack (QSS) method to correct for the first-order nonlinear motion, namely the angular acceleration of cislunar targets. Simulations of typical cislunar orbits verify this theoretical criterion and show that the QSS method significantly improves SNR from stacking and can enhance the detection limit by up to 1 stellar magnitude compared with the linear-motion stacking method. Furthermore, tests using observational data of the cislunar object Tiandu-1 confirm that while linear stacking degrades after a 29-minute integration due to trajectory curvature, the QSS method achieves continuous SNR improvement over a 46-minute integration, outperforming the peak SNR of the linear method by 31%.

astro-ph.IM↗
arXiv2026

Widget2Code: From Visual Widgets to UI Code via Multimodal LLMs

User interface to code (UI2Code) aims to generate executable code that can faithfully reconstruct a given input UI. Prior work focuses largely on web pages and mobile screens, leaving app widgets underexplored. Unlike web or mobile UIs with rich hierarchical context, widgets are compact, context-free micro-interfaces that summarize key information through dense layouts and iconography under strict spatial constraints. Moreover, while (image, code) pairs are widely available for web or mobile UIs, widget designs are proprietary and lack accessible markup. We formalize this setting as the Widget-to-Code (Widget2Code) and introduce an image-only widget benchmark with fine-grained, multi-dimensional evaluation metrics. Benchmarking shows that although generalized multimodal large language models (MLLMs) outperform specialized UI2Code methods, they still produce unreliable and visually inconsistent code. To address these limitations, we develop a baseline that jointly advances perceptual understanding and structured code generation. At the perceptual level, we follow widget design principles to assemble atomic components into complete layouts, equipped with icon retrieval and reusable visualization modules. At the system level, we design an end-to-end infrastructure, WidgetFactory, which includes a framework-agnostic widget-tailored domain-specific language (WidgetDSL) and a compiler that translates it into multiple front-end implementations (e.g., React, HTML/CSS). An adaptive rendering module further refines spatial dimensions to satisfy compactness constraints. Together, these contributions substantially enhance visual fidelity, establishing a strong baseline and unified infrastructure for future Widget2Code research.

cs.CV↗
arXiv2026

Type-II Antiferroelectricity

Antiferroelectricity (AFE) is a fundamental concept in physics and materials science. Conventional AFEs have the picture of alternating local electric dipoles defined in real space. Here, we discover a new class of AFEs, termed type-II AFEs, which possess opposite polarizations defined in momentum space across a pair of symmetry decoupled subspaces. Unlike conventional AFEs, the order parameter of type-II AFEs is rigorously formulated through Berry-phase theory and can be quantitatively extracted from the electronic band structure. Focusing on a subclass of type-II AFEs that preserve spin-rotation symmetry, we establish the relevant symmetry constraints and identify all compatible spin point groups. Remarkably, we find that type-II AFE order intrinsically coexists with antiferromagnetism, revealing a robust form of magnetoelectric coupling. We construct an altermagnetic model and identify several concrete antiferromagnetic/altermagnetic materials, such as FeS, Cr2O3, MgMnO3, monolayer MoICl2 and bilayer CrI3, that exhibit this novel ordering. Furthermore, we uncover unique physical phenomena associated with type-II spin-AFE systems, including spin current generation upon AFE switching and localized spin polarization at boundaries and domain walls. Our findings reveal a previously hidden class of quantum materials with intertwined ferroic orders, offering exciting opportunities for both fundamental exploration and technological applications.

cond-mat.mtrl-sci↗
arXiv2026

LUT-LLM: Efficient Large Language Model Inference with Memory-based Computations on FPGAs

The rapid development of large language models (LLM) has greatly enhanced everyday applications. While many FPGA-based accelerators, with flexibility for fine-grained data control, exhibit superior speed and energy efficiency compared to GPUs, recent GPU-specific optimizations have diminished this advantage. When limited to arithmetic-based computation, FPGAs often underperform GPUs due to their comparatively fewer computational resources. To address this challenge, we exploit a key advantage of FPGAs over GPUs: abundant distributed on-chip memory embedded among computational units. We believe that shifting LLM inference from arithmetic-based to memory-based computations through table lookups can improve the efficiency on FPGAs to compete with GPUs. However, existing methods are inefficient or unable to scale and deploy language models due to algorithm and architecture design limitations. This paper introduces \textbf{LUT-LLM}, the first FPGA accelerator that deploy 1B+ language model with memory-based computation, leveraging vector quantization. We construct a performance model, evaluate multiple quantization schemes, and identify activation-weight vector co-quantization as the most effective approach. To support this scheme, LUT-LLM features (1) bandwidth-aware parallel centroid search to reduce decoding latency, (2) efficient 2D table lookups, and (3) a spatial-temporal hybrid design to reduce data caching for a higher throughput table lookup. We develop a training recipe that converts existing models to support table lookups with high accuracy and prototype LUT-LLM for Qwen 3 1.7B model on the AMD V80 FPGA, reducing arithmetic operations by $4\times$ and achieving a $1.10\sim3.29\times$ faster generation speed and a $3.05\sim 6.60\times$ higher energy efficiency than GPUs.

cs.AR↗