arXiv subjects

Qiang Zhang

Publications and source records attributed to Qiang Zhang.

At least 181 records · Page 10Linked to original sources

arXiv2025

Advancing biomolecular understanding and design following human instructions

Understanding and designing biomolecules, such as proteins and small molecules, is central to advancing drug discovery, synthetic biology and enzyme engineering. Recent breakthroughs in artificial intelligence have revolutionized biomolecular research, achieving remarkable accuracy in biomolecular prediction and design. However, a critical gap remains between artificial intelligence's computational capabilities and researchers' intuitive goals, particularly in using natural language to bridge complex tasks with human intentions. Large language models have shown potential to interpret human intentions, yet their application to biomolecular research remains nascent due to challenges including specialized knowledge requirements, multimodal data integration, and semantic alignment between natural language and biomolecules. To address these limitations, we present InstructBioMol, a large language model designed to bridge natural language and biomolecules through a comprehensive any-to-any alignment of natural language, molecules and proteins. This model can integrate multimodal biomolecules as the input, and enable researchers to articulate design goals in natural language, providing biomolecular outputs that meet precise biological needs. Experimental results demonstrate that InstructBioMol can understand and design biomolecules following human instructions. In particular, it can generate drug molecules with a 10% improvement in binding affinity and design enzymes that achieve an enzyme-substrate pair prediction score of 70.4. This highlights its potential to transform real-world biomolecular research. The code is available at https://github.com/HICAI-ZJU/InstructBioMol.

cs.CL↗

arXiv2025

Silicon single-photon detector achieving over 84% photon detection efficiency with flexible operation modes

Silicon single-photon detectors (Si SPDs) play a crucial role in detecting single photons in the visible spectrum. For various applications, photon detection efficiency (PDE) is the most critical characteristic for effectively collecting photons. Here, we present a Si SPD with a remarkable PDE of up to 84.4% at 785 nm, supporting multiple operation modes. We design and fabricate a thick-junction Si single-photon avalanche diode (SPAD) that enhances the avalanche probability through a backside-illumination structure, while minimizing noise through the design of a doping-compensated avalanche region. To maximize PDE, we implement a readout circuit with a 50 V quenching voltage, enabling operation in free-running, gating, or hybrid modes. The SPAD, along with its readout circuits and affiliated circuits, is integrated into a compact SPD module. In free-running mode, the module achieves a maximum PDE of 84.4%, with a dark count rate of 260 cps, and an afterpulse probability of 2.9% at 268 K. This work provides a practical solution for applications requiring ultra-high-efficiency Si SPD with multiple operation modes.

quant-ph↗

arXiv2025

Transfer Attack for Bad and Good: Explain and Boost Adversarial Transferability across Multimodal Large Language Models

Multimodal Large Language Models (MLLMs) demonstrate exceptional performance in cross-modality interaction, yet they also suffer adversarial vulnerabilities. In particular, the transferability of adversarial examples remains an ongoing challenge. In this paper, we specifically analyze the manifestation of adversarial transferability among MLLMs and identify the key factors that influence this characteristic. We discover that the transferability of MLLMs exists in cross-LLM scenarios with the same vision encoder and indicate \underline{\textit{two key Factors}} that may influence transferability. We provide two semantic-level data augmentation methods, Adding Image Patch (AIP) and Typography Augment Transferability Method (TATM), which boost the transferability of adversarial examples across MLLMs. To explore the potential impact in the real world, we utilize two tasks that can have both negative and positive societal impacts: \ding{182} Harmful Content Insertion and \ding{183} Information Protection.

cs.CV↗

arXiv2025

Learning Null Geodesics for Gravitational Lensing Rendering in General Relativity

We present GravLensX, an innovative method for rendering black holes with gravitational lensing effects using neural networks. The methodology involves training neural networks to fit the spacetime around black holes and then employing these trained models to generate the path of light rays affected by gravitational lensing. This enables efficient and scalable simulations of black holes with optically thin accretion disks, significantly decreasing the time required for rendering compared to traditional methods. We validate our approach through extensive rendering of multiple black hole systems with superposed Kerr metric, demonstrating its capability to produce accurate visualizations with significantly $15\times$ reduced computational time. Our findings suggest that neural networks offer a promising alternative for rendering complex astrophysical phenomena, potentially paving a new path to astronomical visualization.

gr-qc↗

arXiv2025

Quenched Disorder in the Triangular Lattice Antiferromagnet YbZn$_2$GaO$_5$

We investigate the crystal electric field (CEF) excitations of Yb$^{3+}$ ions in powder samples of the triangular-lattice rare-earth-based antiferromagnet YbZn$_2$GaO$_5$ using inelastic neutron scattering (INS). Three CEF excitations from the ground-state Kramers doublet were observed, each exhibiting significant broadening beyond instrumental resolution. Combining temperature-dependent INS and neutron powder diffraction, we identify a significant static contribution to this broadening and attribute it to heterogeneous coordination of Yb$^{3+}$ ions due to Ga$^{3+}$/Zn$^{2+}$ site mixing. Rietveld refinement of neutron powder diffraction indicates that 35% of Ga occupies the Zn site and 60% of Zn occupies the Ga site. We show with a point charge model for the CEF Hamiltonian that heterogeneous coordination of Yb$^{3+}$ ions leads to broadened CEF peaks. First-principles calculations demonstrate that the random Ga$^{3+}$/Zn$^{2+}$ distribution can produce the distortions of the YbO$_6$ octahedra observed from neutron diffraction. Because the documented heterogeneity will extend to exchange interactions, our results suggest that disorder is a significant factor in the unusual magnetism previously reported in YbZn$_2$GaO$_5$, including broad low-energy magnetic excitations and the absence of magnetic ordering down to 0.3K.

cond-mat.str-el↗

arXiv2025

Enhancing Safe and Controllable Protein Generation via Knowledge Preference Optimization

Protein language models have emerged as powerful tools for sequence generation, offering substantial advantages in functional optimization and denovo design. However, these models also present significant risks of generating harmful protein sequences, such as those that enhance viral transmissibility or evade immune responses. These concerns underscore critical biosafety and ethical challenges. To address these issues, we propose a Knowledge-guided Preference Optimization (KPO) framework that integrates prior knowledge via a Protein Safety Knowledge Graph. This framework utilizes an efficient graph pruning strategy to identify preferred sequences and employs reinforcement learning to minimize the risk of generating harmful proteins. Experimental results demonstrate that KPO effectively reduces the likelihood of producing hazardous sequences while maintaining high functionality, offering a robust safety assurance framework for applying generative models in biotechnology.

cs.AI↗

arXiv2025

E-21 Level Instability Frequency Dissemination over 2067 km noisy Telecommunication Infrastructure

The realization of ultra stable optical frequency transmission through fiber networks is critical for advancing global optical frequency standards and enabling applications such as redefining the second in the International System of Units, geophysical sensing, quantum network construction, and fundamental physics experiments. However, achieving high reliability and low instability optical frequency carrier transmission links over distances exceeding thousands of kilometers remains technically challenging, thereby limiting the scalability and reliability of such networks. In this study, we experimentally demonstrate that the noise accumulation in long distance optical links can be mitigated by narrowband purification of the optical signal's phase noise, enabling optical links of theoretically unlimited length. Additionally, we implemented digital optical phase measurement and feedback technology to calibrate noise compensation deviations caused by inconsistencies in round trip optical frequencies, enhancing link stability. By adopting digital phase measurement instead of traditional phase detectors, we expanded the dynamic noise tolerance range of the optical phase-locked loop, significantly improving system reliability. Ultimately, on a 2067 km telecommunications fiber link with a noise level exceeding 5000 rad^2/Hz.km, we achieved an optical frequency transfer with a daily instability of 2.9 E-21 without experiencing any optical cycle slips maintaining continuous operation for four days. This work establishes a technical foundation for leveraging existing fiber resources to construct global scale optical frequency standard networks.

physics.optics↗

arXiv2025

Constrain Alignment with Sparse Autoencoders

The alignment of large language models (LLMs) with human preferences remains a key challenge. While post-training techniques like Reinforcement Learning from Human Feedback (RLHF) and Direct Preference Optimization (DPO) have achieved notable success, they often introduce computational inefficiencies and training instability. In this paper, we propose Feature-level constrained Preference Optimization (FPO), a novel method designed to simplify the alignment process while ensuring stability. FPO leverages pre-trained Sparse Autoencoders (SAEs) and introduces feature-level constraints, allowing for efficient, sparsity-enforced alignment. Our approach enjoys efficiency by using sparse features activated in a well-trained sparse autoencoder and the quality of sequential KL divergence by using the feature-level offline reference. Experimental results on benchmark datasets demonstrate that FPO achieves a 5.08% absolute improvement in win rate with much lower computational cost compared to state-of-the-art baselines, making it a promising solution for efficient and controllable LLM alignments.

cs.AI↗

arXiv2025

LOVON: Legged Open-Vocabulary Object Navigator

Object navigation in open-world environments remains a formidable and pervasive challenge for robotic systems, particularly when it comes to executing long-horizon tasks that require both open-world object detection and high-level task planning. Traditional methods often struggle to integrate these components effectively, and this limits their capability to deal with complex, long-range navigation missions. In this paper, we propose LOVON, a novel framework that integrates large language models (LLMs) for hierarchical task planning with open-vocabulary visual detection models, tailored for effective long-range object navigation in dynamic, unstructured environments. To tackle real-world challenges including visual jittering, blind zones, and temporary target loss, we design dedicated solutions such as Laplacian Variance Filtering for visual stabilization. We also develop a functional execution logic for the robot that guarantees LOVON's capabilities in autonomous navigation, task adaptation, and robust task completion. Extensive evaluations demonstrate the successful completion of long-sequence tasks involving real-time detection, search, and navigation toward open-vocabulary dynamic targets. Furthermore, real-world experiments across different legged robots (Unitree Go2, B2, and H1-2) showcase the compatibility and appealing plug-and-play feature of LOVON.

cs.RO↗

arXiv2025

Boosting LLM's Molecular Structure Elucidation with Knowledge Enhanced Tree Search Reasoning

Molecular structure elucidation involves deducing a molecule's structure from various types of spectral data, which is crucial in chemical experimental analysis. While large language models (LLMs) have shown remarkable proficiency in analyzing and reasoning through complex tasks, they still encounter substantial challenges in molecular structure elucidation. We identify that these challenges largely stem from LLMs' limited grasp of specialized chemical knowledge. In this work, we introduce a Knowledge-enhanced reasoning framework for Molecular Structure Elucidation (K-MSE), leveraging Monte Carlo Tree Search for test-time scaling as a plugin. Specifically, we construct an external molecular substructure knowledge base to extend the LLMs' coverage of the chemical structure space. Furthermore, we design a specialized molecule-spectrum scorer to act as a reward model for the reasoning process, addressing the issue of inaccurate solution evaluation in LLMs. Experimental results show that our approach significantly boosts performance, particularly gaining more than 20% improvement on both GPT-4o-mini and GPT-4o. Our code is available at https://github.com/HICAI-ZJU/K-MSE.

cs.CL↗

arXiv2025

A deep-learning model for predicting daily PM2.5 concentration in response to emission reduction

Air pollution remains a leading global health threat, with fine particulate matter (PM2.5) contributing to millions of premature deaths annually. Chemical transport models (CTMs) are essential tools for evaluating how emission controls improve air quality and save lives, but they are computationally intensive. Reduced form models accelerate simulations but sacrifice spatial-temporal granularity, accuracy, and flexibility. Here we present CleanAir, a deep-learning-based model developed as an efficient alternative to CTMs in simulating daily PM2.5 and its chemical compositions in response to precursor emission reductions at 36 km resolution, which could predict PM2.5 concentration for a full year within 10 seconds on a single GPU, a speed five orders of magnitude faster. Built on a Residual Symmetric 3D U-Net architecture and trained on more than 2,400 emission reduction scenarios generated by a well-validated Community Multiscale Air Quality (CMAQ) model, CleanAir generalizes well across unseen meteorological years and emission patterns. It produces results comparable to CMAQ in both absolute concentrations and emission-induced changes, enabling efficient, full-coverage simulations across short-term interventions and long-term planning horizons. This advance empowers researchers and policymakers to rapidly evaluate a wide range of air quality strategies and assess the associated health impacts, thereby supporting more responsive and informed environmental decision-making.

physics.ao-ph↗

arXiv2025

RiOT: Efficient Prompt Refinement with Residual Optimization Tree

Recent advancements in large language models (LLMs) have highlighted their potential across a variety of tasks, but their performance still heavily relies on the design of effective prompts. Existing methods for automatic prompt optimization face two challenges: lack of diversity, limiting the exploration of valuable and innovative directions and semantic drift, where optimizations for one task can degrade performance in others. To address these issues, we propose Residual Optimization Tree (RiOT), a novel framework for automatic prompt optimization. RiOT iteratively refines prompts through text gradients, generating multiple semantically diverse candidates at each step, and selects the best prompt using perplexity. Additionally, RiOT incorporates the text residual connection to mitigate semantic drift by selectively retaining beneficial content across optimization iterations. A tree structure efficiently manages the optimization process, ensuring scalability and flexibility. Extensive experiments across five benchmarks, covering commonsense, mathematical, logical, temporal, and semantic reasoning, demonstrate that RiOT outperforms both previous prompt optimization methods and manual prompting.

cs.CL↗

arXiv2025

EFormer: An Effective Edge-based Transformer for Vehicle Routing Problems

Recent neural heuristics for the Vehicle Routing Problem (VRP) primarily rely on node coordinates as input, which may be less effective in practical scenarios where real cost metrics-such as edge-based distances-are more relevant. To address this limitation, we introduce EFormer, an Edge-based Transformer model that uses edge as the sole input for VRPs. Our approach employs a precoder module with a mixed-score attention mechanism to convert edge information into temporary node embeddings. We also present a parallel encoding strategy characterized by a graph encoder and a node encoder, each responsible for processing graph and node embeddings in distinct feature spaces, respectively. This design yields a more comprehensive representation of the global relationships among edges. In the decoding phase, parallel context embedding and multi-query integration are used to compute separate attention mechanisms over the two encoded embeddings, facilitating efficient path construction. We train EFormer using reinforcement learning in an autoregressive manner. Extensive experiments on the Traveling Salesman Problem (TSP) and Capacitated Vehicle Routing Problem (CVRP) reveal that EFormer outperforms established baselines on synthetic datasets, including large-scale and diverse distributions. Moreover, EFormer demonstrates strong generalization on real-world instances from TSPLib and CVRPLib. These findings confirm the effectiveness of EFormer's core design in solving VRPs.

cs.LG↗

arXiv2025

Ultrabright Entanglement Based Quantum Key Distribution over a 404 km Optical Fiber

Entangled photons are crucial resources for quantum information processing. Here, we present an ultrabright polarization-entangled photon source based on a periodically poled lithium niobate waveguide designed for practical quantum communication networks. Using a 780 nm pump laser, the source achieves a pair generation rate of 2.4 $\times 10^{10}$ pairs/s/mW. Remarkably, the entangled photons are bright enough to be detected by a power meter, reaching a power of 17.9 nW under a pump power of 3.2 mW. We demonstrate the practicality of the source by conducting quantum key distribution experiments over long-distance fiber links. Wavelength-division multiplexing was employed to enhance the key generation, and nonlocal dispersion compensation was implemented to ensure precise timing coincidence measurements across a broad spectral range. By utilizing nine pairs of wavelength channels, the system achieved the applicable secure key rates of up to 440.80 bits/s over 200 km with a 62 dB loss and extended the maximum secure key generation distance to 404 km. These results demonstrate the potential of wavelength-multiplexed polarization-entangled photon sources for high-speed, long-distance quantum communication, positioning them as key components for future large-scale quantum networks.

quant-ph↗

arXiv2025

Meta-heuristic Hypergraph-Assisted Robustness Optimization for Higher-order Complex Systems

In complex systems (e.g., communication, transportation, and biological networks), high robustness ensures sustained functionality and stability even when resisting attacks. However, the inherent structure complexity and the unpredictability of attacks make robustness optimization challenging. Hypergraphs provide a framework for modeling complicated higher-order interactions in complex systems naturally, but their potential has not been systematically investigated. Therefore, we propose an effective method based on genetic algorithms from Artificial Intelligence to optimize the robustness of complex systems modeled by hypergraphs. By integrating percolation-based metrics with adaptive computational techniques, our method achieves improved accuracy and efficiency. Experiments on both synthetic and real-world hypergraphs demonstrate the effectiveness of the proposed method in mitigating malicious attacks, with robustness improvements ranging from 16.6% to 205.2%. Further in-depth analysis reveals that optimized hypergraph-based systems exhibit a preferential connection mechanism in which high-hyperdegree nodes preferentially connect to lower-cardinality hyperedges, forming a distinctive Lotus topology that significantly improves robustness. Based on this finding, we propose a robust hypergraph generation method that allows robustness to be controlled via a single parameter rb. Notably, for rb<-1, a distinct Cactus topology emerges as an alternative to the Lotus topology observed for rb>1. The discovery of the Lotus and Cactus topologies offers valuable insights for designing robust higher-order networks while providing a useful foundation for investigating cascading failure dynamics in complex systems.

physics.soc-ph↗

arXiv2025

GEM: Empowering LLM for both Embedding Generation and Language Understanding

Large decoder-only language models (LLMs) have achieved remarkable success in generation and reasoning tasks, where they generate text responses given instructions. However, many applications, e.g., retrieval augmented generation (RAG), still rely on separate embedding models to generate text embeddings, which can complicate the system and introduce discrepancies in understanding of the query between the embedding model and LLMs. To address this limitation, we propose a simple self-supervised approach, Generative Embedding large language Model (GEM), that enables any large decoder-only LLM to generate high-quality text embeddings while maintaining its original text generation and reasoning capabilities. Our method inserts new special token(s) into a text body, and generates summarization embedding of the text by manipulating the attention mask. This method could be easily integrated into post-training or fine tuning stages of any existing LLMs. We demonstrate the effectiveness of our approach by applying it to two popular LLM families, ranging from 1B to 8B parameters, and evaluating the transformed models on both text embedding benchmarks (MTEB) and NLP benchmarks (MMLU). The results show that our proposed method significantly improves the original LLMs on MTEB while having a minimal impact on MMLU. Our strong results indicate that our approach can empower LLMs with state-of-the-art text embedding capabilities while maintaining their original NLP performance

cs.CL↗

arXiv2025

Classification of aut-fixed subgroups in free-abelian times surface groups

In this paper, we are concerned with the direct product $G=π_1(Σ)\times \Z^k$ for $Σ$ a compact orientable surface with negative Euler characteristic, and give a complete classification of its fixed subgroups of automorphisms. As a corollary, we show that $G$ contains, up to isomorphism, infinitely many fixed subgroups of automorphisms if and only if $k\geq 2$, which is a contrast to that of hyperbolic groups. As an application on Nielsen fixed point theory, we provide a family of aspherical manifolds without Jiang's Bound Index Property. Moreover, we also give some results on the fixed subgroups of the direct product $H\times \Z^k$ for $H$ a non-elementary torsion-free hyperbolic group.

math.GR↗

arXiv2025

Sample-Efficient Human Evaluation of Large Language Models via Maximum Discrepancy Competition

Reliable evaluation of large language models (LLMs) is impeded by two key challenges: objective metrics often fail to reflect human perception of natural language, and exhaustive human labeling is prohibitively expensive. Here, we propose a sample-efficient human evaluation method for LLMs based on the principle of MAximum Discrepancy (MAD) Competition. Our method automatically and adaptively selects a compact set of input instructions that maximize semantic discrepancy between pairs of LLM responses. Human evaluators then perform three-alternative forced choices on these paired responses, which are aggregated into a global ranking using Elo rating. We apply our approach to compare eight widely used LLMs across four tasks: scientific knowledge understanding, mathematical reasoning, creative and functional writing, and code generation and explanation. Experimental results show that our sample-efficient evaluation method recovers "gold-standard" model rankings with a handful of MAD-selected instructions, reveals respective strengths and weaknesses of each LLM, and offers nuanced insights to guide future LLM development. Code is available at https://github.com/weiji-Feng/MAD-Eval .

cs.LG↗