arXiv subjects

Haoran Wang

Publications and source records attributed to Haoran Wang.

At least 109 records · Page 6Linked to original sources

arXiv2025

OpenRLHF: An Easy-to-use, Scalable and High-performance RLHF Framework

Large Language Models (LLMs) fine-tuned via Reinforcement Learning from Human Feedback (RLHF) and Reinforcement Learning with Verifiable Rewards (RLVR) significantly improve the alignment of human-AI values, further raising the upper bound of AI capabilities, particularly in reasoning-intensive, long-context Chain-of-Thought (CoT) tasks. However, existing frameworks commonly face challenges such as inference bottlenecks and complexity barriers, which restrict their accessibility to newcomers. To bridge this gap, we introduce \textbf{OpenRLHF}, a user-friendly, scalable, and easy-to-learn open-source RLHF framework built upon Ray, vLLM, DeepSpeed, and HuggingFace Transformers, featuring a simplified design, clear code structure, and comprehensive documentation to facilitate entry for researchers and practitioners. Experimental results show that OpenRLHF achieves superior training efficiency, with speedups ranging from 1.22x to 1.68x across different model sizes, compared to state-of-the-art frameworks. Additionally, it requires significantly fewer lines of code for implementation. OpenRLHF is publicly available at https://github.com/OpenRLHF/OpenRLHF, and has already been adopted by leading institutions to accelerate RLHF research and learning.

cs.AI↗

arXiv2025

The land use-climate change-biodiversity nexus in European islands stakeholders

To promote climate adaptation and mitigation, it is crucial to understand stakeholder perspectives and knowledge gaps on land use and climate changes. Stakeholders across 21 European islands were consulted on climate and land use change issues affecting ecosystem services. Climate change perceptions included temperature, precipitation, humidity, extremes, and wind. Land use change perceptions included deforestation, coastal degradation, habitat protection, renewable energy facilities, wetlands, and others. Additional concerns such as invasive species, water or energy scarcity, infrastructure problems, and austerity were also considered. Climate and land use change impact perceptions were analysed with machine learning to quantify their influence. The predominant climatic characteristic is temperature, and the predominant land use characteristic is deforestation. Water-related problems are top priorities for stakeholders. Energy-related problems, including energy deficiency and issues with wind and solar facilities, rank high as combined climate and land use risks. Stakeholders generally perceive climate change impacts on ecosystem services as negative, with natural habitat destruction and biodiversity loss identified as top issues. Land use change impacts are also negative but more complex, with more explanatory variables. Stakeholders share common perceptions on biodiversity impacts despite geographic disparity, but they differentiate between climate and land use impacts. Water, energy, and renewable energy issues pose serious concerns, requiring management measures.

q-bio.PE↗

arXiv2025

Integrated Silicon Photonic Multichannel Optical Hybrid for Broadband Parallel Coherent Reception

We design and demonstrate a monolithically integrated silicon photonic multichannel optical hybrid for versatile broadband coherent reception, addressing the critical limitations of current wavelength multiplexed systems in scalability and power efficiency. The device combines a phase-compensated 90-degree optical hybrid with four robust three-stage Mach-Zehnder interferometer lattice filters, enabling 34-port functionality (two inputs and 32 outputs) for simultaneous analog and digital signal processing. Leveraging multimode interferometer designs,the chip achieves a broadband response with sub-dB passband uniformity across eight 200 GHz-spaced wavelength channels, while maintaining phase errors below 4 degrees over a 13.5 nm (1539-1552.5 nm) bandwidth with only 2.5 mW thermal tuning power.Experimentally, we validate its parallel-processing capability through RF channelizer reception (showing an average spurious-free dynamic range of 80.8 dB*Hz2/3 and image rejection ratio of 33.26 dB) and coherent optical communication (achieving 1.024 Tb/s data rate for 32-QAM signals with bit error rates far below the 20% SD-FEC threshold). The scheme enhances system performance with fully passive wavelength multiplexing integration, supporting high-fidelity uniformity and projecting scalability to 1.468 Tb/s. This work promises advancements in high-performance optoelectronic devices for next-generation AI-driven data centers and 5G-XG networks.

physics.optics↗

arXiv2025

From Restoration to Reconstruction: Rethinking 3D Gaussian Splatting for Underwater Scenes

Underwater image degradation poses significant challenges for 3D reconstruction, where simplified physical models often fail in complex scenes. We propose \textbf{R-Splatting}, a unified framework that bridges underwater image restoration (UIR) with 3D Gaussian Splatting (3DGS) to improve both rendering quality and geometric fidelity. Our method integrates multiple enhanced views produced by diverse UIR models into a single reconstruction pipeline. During inference, a lightweight illumination generator samples latent codes to support diverse yet coherent renderings, while a contrastive loss ensures disentangled and stable illumination representations. Furthermore, we propose \textit{Uncertainty-Aware Opacity Optimization (UAOO)}, which models opacity as a stochastic function to regularize training. This suppresses abrupt gradient responses triggered by illumination variation and mitigates overfitting to noisy or view-specific artifacts. Experiments on Seathru-NeRF and our new BlueCoral3D dataset demonstrate that R-Splatting outperforms strong baselines in both rendering quality and geometric accuracy.

cs.CV↗

arXiv2025

Qianfan-VL: Domain-Enhanced Universal Vision-Language Models

We present Qianfan-VL, a series of multimodal large language models ranging from 3B to 70B parameters, achieving state-of-the-art performance through innovative domain enhancement techniques. Our approach employs multi-stage progressive training and high-precision data synthesis pipelines, which prove to be critical technologies for enhancing domain-specific capabilities while maintaining strong general performance. Qianfan-VL achieves comparable results to leading open-source models on general benchmarks, with state-of-the-art performance on benchmarks such as CCBench, SEEDBench IMG, ScienceQA, and MMStar. The domain enhancement strategy delivers significant advantages in OCR and document understanding, validated on both public benchmarks (OCRBench 873, DocVQA 94.75%) and in-house evaluations. Notably, Qianfan-VL-8B and 70B variants incorporate long chain-of-thought capabilities, demonstrating superior performance on mathematical reasoning (MathVista 78.6%) and logical inference tasks. All models are trained entirely on Baidu's Kunlun P800 chips, validating the capability of large-scale AI infrastructure to train SOTA-level multimodal models with over 90% scaling efficiency on 5000 chips for a single task. This work establishes an effective methodology for developing domain-enhanced multimodal models suitable for diverse enterprise deployment scenarios.

cs.CV↗

arXiv2025

Versatile and reconfigurable integrated silicon nitride photonic microresonator

Unlocking the full potential of integrated photonics requires versatile, multi-functional devices that can adapt to diverse application demands. However, confronting this challenge with conventional single-function resonators often results in tedious and complex systems. We present an elegant solution: a versatile and reconfigurable dual-polarization Si3N4 microresonator that represents a paradigm shift in on-chip photonic designs. Our device, based on a binary-star orbital architecture, can be dynamically reconfigured into three distinct topologies: a Möbius-like microcavity, a Fabry-Pérot resonator, and a microring resonator. This unprecedented functionality is enabled by a tunable balanced Mach-Zehnder interferometer that facilitates controllable mutual mode coupling of counterpropagating lights using a single control knob. We experimentally demonstrate that the device not only supports polarization-diverse operation on a compact footprint but also gives rise to a rich variety of physical phenomena, including a standing wave cavity, a traveling wave cavity, free spectral range multiplication, and the photonic pinning effect. These behaviors are accurately modeled using the Transfer Matrix Method and intuitively explained by Temporal Coupled Mode Theory. Our results underscore the profound potential for a chip-scale platform to realize reconfigurable reconstructive spectrometers and on-chip synthetic dimensions for topological physics.

physics.optics↗

arXiv2025

Correspondence as Video: Test-Time Adaption on SAM2 for Reference Segmentation in the Wild

Large vision models like the Segment Anything Model (SAM) exhibit significant limitations when applied to downstream tasks in the wild. Consequently, reference segmentation, which leverages reference images and their corresponding masks to impart novel knowledge to the model, emerges as a promising new direction for adapting vision models. However, existing reference segmentation approaches predominantly rely on meta-learning, which still necessitates an extensive meta-training process and brings massive data and computational cost. In this study, we propose a novel approach by representing the inherent correspondence between reference-target image pairs as a pseudo video. This perspective allows the latest version of SAM, known as SAM2, which is equipped with interactive video object segmentation (iVOS) capabilities, to be adapted to downstream tasks in a lightweight manner. We term this approach Correspondence As Video for SAM (CAV-SAM). CAV-SAM comprises two key modules: the Diffusion-Based Semantic Transition (DBST) module employs a diffusion model to construct a semantic transformation sequence, while the Test-Time Geometric Alignment (TTGA) module aligns the geometric changes within this sequence through test-time fine-tuning. We evaluated CAVSAM on widely-used datasets, achieving segmentation performance improvements exceeding 5% over SOTA methods. Implementation is provided in the supplementary materials.

cs.CV↗

arXiv2025

GLM-4.5: Agentic, Reasoning, and Coding (ARC) Foundation Models

We present GLM-4.5, an open-source Mixture-of-Experts (MoE) large language model with 355B total parameters and 32B activated parameters, featuring a hybrid reasoning method that supports both thinking and direct response modes. Through multi-stage training on 23T tokens and comprehensive post-training with expert model iteration and reinforcement learning, GLM-4.5 achieves strong performance across agentic, reasoning, and coding (ARC) tasks, scoring 70.1% on TAU-Bench, 91.0% on AIME 24, and 64.2% on SWE-bench Verified. With much fewer parameters than several competitors, GLM-4.5 ranks 3rd overall among all evaluated models and 2nd on agentic benchmarks. We release both GLM-4.5 (355B parameters) and a compact version, GLM-4.5-Air (106B parameters), to advance research in reasoning and agentic AI systems. Code, models, and more information are available at https://github.com/zai-org/GLM-4.5.

cs.CL↗

arXiv2025

Human-Machine Shared Control Approach for the Takeover of CACC

Cooperative Adaptive Cruise Control (CACC) often requires human takeover for tasks such as exiting a freeway. Direct human takeover can pose significant risks, especially given the close-following strategy employed by CACC, which might cause drivers to feel unsafe and execute hard braking, potentially leading to collisions. This research aims to develop a CACC takeover controller that ensures a smooth transition from automated to human control. The proposed CACC takeover maneuver employs an indirect human-machine shared control approach, modeled as a Stackelberg competition where the machine acts as the leader and the human as the follower. The machine guides the human to respond in a manner that aligns with the machine's expectations, aiding in maintaining following stability. Additionally, the human reaction function is integrated into the machine's predictive control system, moving beyond a simple "prediction-planning" pipeline to enhance planning optimality. The controller has been verified to i) enable a smooth takeover maneuver of CACC; ii) ensure string stability in the condition that the platoon has less than 6 CAVs and human control authority is less than 40%; iii) enhance both perceived and actual safety through machine interventions; and iv) reduce the impact on upstream traffic by up to 60%.

cs.RO↗

arXiv2025

Manipulating non-abelian anyons of Kitaev quantum spin liquids with local magnetic fields

Non-abelian anyonic excitations of quantum spin liquids have potential for application to topological quantum computation, but designing logical operations requires developing protocols to faithfully create, move, and read-out such quasiparticles. In this paper, we present a protocol that manipulates the $Z_2$ fluxes (``visons''), and their bound Majorana zero mode, of the Kitaev model. This is achieved by adiabatically switching-off and -on interaction terms in the Hamiltonian, and applying a local magnetic field. We test our protocol by exchanging two Majorana zero modes, analysing the errors it produces. We find that the error rate of our protocol can be suppressed exponentially by increasing the adiabatic turn-on time and the size of the system. However, realistic implementations must consider the trade off between different error sources.

cond-mat.str-el↗

arXiv2025

Extreme Cardiac MRI Analysis under Respiratory Motion: Results of the CMRxMotion Challenge

Deep learning models have achieved state-of-the-art performance in automated Cardiac Magnetic Resonance (CMR) analysis. However, the efficacy of these models is highly dependent on the availability of high-quality, artifact-free images. In clinical practice, CMR acquisitions are frequently degraded by respiratory motion, yet the robustness of deep learning models against such artifacts remains an underexplored problem. To promote research in this domain, we organized the MICCAI CMRxMotion challenge. We curated and publicly released a dataset of 320 CMR cine series from 40 healthy volunteers who performed specific breathing protocols to induce a controlled spectrum of motion artifacts. The challenge comprised two tasks: 1) automated image quality assessment to classify images based on motion severity, and 2) robust myocardial segmentation in the presence of motion artifacts. A total of 22 algorithms were submitted and evaluated on the two designated tasks. This paper presents a comprehensive overview of the challenge design and dataset, reports the evaluation results for the top-performing methods, and further investigates the impact of motion artifacts on five clinically relevant biomarkers. All resources and code are publicly available at: https://github.com/CMRxMotion

eess.IV↗

arXiv2025

Robust and Efficient Autoregressive Speech Synthesis with Dynamic Chunk-wise Prediction Policy

Recently, autoregressive (AR) language models have emerged as a dominant approach in speech synthesis, offering expressive generation and scalable training. However, conventional AR speech synthesis models relying on the next-token prediction paradigm often encounter significant challenges when handling long speech sequences. These models often struggle to construct stable frame-to-frame attention, leading to increased latency and degraded synthesis quality, thereby limiting their feasibility for real-time applications. To address these limitations, we introduce a novel dynamic chunk-wise autoregressive synthesis framework, termed DCAR, designed to enhance both efficiency and intelligibility robustness in AR speech generation. DCAR introduces a chunk-to-frame attention mechanism through training with multi-token prediction, enabling dynamic chunk prediction in variable speech contexts using a lightweight module trained on-policy. DCAR dynamically adjusts the token prediction span, significantly reducing the sequence length dependency while obtaining high synthesis quality. Comprehensive empirical evaluations demonstrate that DCAR substantially outperforms traditional next-token prediction models, achieving up to 72.27% intelligibility improvement and 2.61x inference speedup simultaneously on the test set. Furthermore, we conduct comprehensive analysis to support it as a versatile foundation for next-generation speech synthesis systems.

cs.SD↗

arXiv2025

MIRAGE: A Multi-modal Benchmark for Spatial Perception, Reasoning, and Intelligence

Spatial perception and reasoning are core components of human cognition, encompassing object recognition, spatial relational understanding, and dynamic reasoning. Despite progress in computer vision, existing benchmarks reveal significant gaps in models' abilities to accurately recognize object attributes and reason about spatial relationships, both essential for dynamic reasoning. To address these limitations, we propose MIRAGE, a multi-modal benchmark designed to evaluate models' capabilities in Counting (object attribute recognition), Relation (spatial relational reasoning), and Counting with Relation. Through diverse and complex scenarios requiring fine-grained recognition and reasoning, MIRAGE highlights critical limitations in state-of-the-art models, underscoring the need for improved representations and reasoning frameworks. By targeting these foundational abilities, MIRAGE provides a pathway toward spatiotemporal reasoning in future research.

cs.CV↗

arXiv2025

SWE-Dev: Building Software Engineering Agents with Training and Inference Scaling

Large language models (LLMs) have advanced rapidly from conversational problem solving to addressing real-world tasks involving tool use, such as software engineering (SWE). Recent LLM-powered toolkits, such as OpenAI Codex and Cursor, have offered end-to-end automation of the software development process. However, building effective SWE agents remains challenging due to the lack of high-quality training data and effective test cases. To address this issue, we present SWE-Dev, an SWE agent built upon open-source LLMs. First, we develop a robust pipeline to synthesize test cases for patch evaluation. Second, we scale up agent trajectories to construct the training data for building SWE-Dev. Experiments on the SWE-bench-Verified benchmark show that the SWE-Dev models can achieve top performance among all open SWE agents. Specifically, the success rates of the SWE-Dev 7B and 32B parameter models reach 23.4% and 36.6%, respectively, outperforming state-of-the-art open-source models. All code, models, and datasets are publicly available at https://github.com/THUDM/SWE-Dev.

cs.AI↗

arXiv2025

RealDriveSim: A Realistic Multi-Modal Multi-Task Synthetic Dataset for Autonomous Driving

As perception models continue to develop, the need for large-scale datasets increases. However, data annotation remains far too expensive to effectively scale and meet the demand. Synthetic datasets provide a solution to boost model performance with substantially reduced costs. However, current synthetic datasets remain limited in their scope, realism, and are designed for specific tasks and applications. In this work, we present RealDriveSim, a realistic multi-modal synthetic dataset for autonomous driving that not only supports popular 2D computer vision applications but also their LiDAR counterparts, providing fine-grained annotations for up to 64 classes. We extensively evaluate our dataset for a wide range of applications and domains, demonstrating state-of-the-art results compared to existing synthetic benchmarks. The dataset is publicly available at https://realdrivesim.github.io/.

cs.CV↗

arXiv2025

E2MPL:An Enduring and Efficient Meta Prompt Learning Framework for Few-shot Unsupervised Domain Adaptation

Few-shot unsupervised domain adaptation (FS-UDA) leverages a limited amount of labeled data from a source domain to enable accurate classification in an unlabeled target domain. Despite recent advancements, current approaches of FS-UDA continue to confront a major challenge: models often demonstrate instability when adapted to new FS-UDA tasks and necessitate considerable time investment. To address these challenges, we put forward a novel framework called Enduring and Efficient Meta-Prompt Learning (E2MPL) for FS-UDA. Within this framework, we utilize the pre-trained CLIP model as the backbone of feature learning. Firstly, we design domain-shared prompts, consisting of virtual tokens, which primarily capture meta-knowledge from a wide range of meta-tasks to mitigate the domain gaps. Secondly, we develop a task prompt learning network that adaptively learns task-specific specific prompts with the goal of achieving fast and stable task generalization. Thirdly, we formulate the meta-prompt learning process as a bilevel optimization problem, consisting of (outer) meta-prompt learner and (inner) task-specific classifier and domain adapter. Also, the inner objective of each meta-task has the closed-form solution, which enables efficient prompt learning and adaptation to new tasks in a single step. Extensive experimental studies demonstrate the promising performance of our framework in a domain adaptation benchmark dataset DomainNet. Compared with state-of-the-art methods, our method has improved accuracy by at least 15.4% and reduced the time by 68.5% on average in 5-way 1-shot tasks, and improved accuracy by 8.7% and reduced the time by 74.1% on average in 5-way 5-shot tasks. Moreover, our approach exhibits more enduring performance than the other methods, i.e., being more stable across 3600 test tasks.

cs.CV↗

arXiv2025

Balancing Efficiency and Equity in Classroom Assignment under Endogenous Peer Effects

This paper presents a three-step empirical framework for optimizing classroom assignments under endogenous peer effects, using data from the China Education Panel Survey (CEPS). We design \textit{PeerNN}, a neural network that mimics endogenous network formation as a discrete choice model, generating a friendship-intensity matrix ($Ω$) that captures student popularity. \textbf{Step 2: Estimating Peer Effects.} We measure the peer effect friends' average 6th-grade class rank weighted by $Ω$ on 8th-grade cognitive test score. Incorporating $Ω$ into the linear-in-means model induces endogeneity. Using quasi-random classroom assignments, we instrument friends' average 6th-grade class rank with the average classmates' 6th-grade class rank (unweighted by $Ω$). Our main regression result shows that a 10\% improvement in friends' 6th-grade class rank raises 8th-grade cognitive test scores by 0.13 SD. Positive $β$ implies maximizing (minimizing) the popularity of high (low) achievers optimizes outcomes. \textbf{Step 3: Simulating Policy Trade-offs.} We use estimates from Step 1 and Step 2 to simulate optimal classroom assignments. We first implement a genetic algorithm (GA) to maximize average peer effect and observe a 1.9\% improvement. However, serious inequity issues arise: low-achieving students are hurt the most in the pursuit of the higher average peer effect. We propose an \textit{Algorithmically Fair GA} (AFGA), achieving a 1.2\% gain while ensuring more equitable educational outcomes. These results underscore that efficiency-focused classroom assignment policies can exacerbate inequality. We recommend incorporating fairness considerations when designing classroom assignment policies that account for endogenous spillovers.

econ.GN↗

arXiv2025

RenderBender: A Survey on Adversarial Attacks Using Differentiable Rendering

Differentiable rendering techniques like Gaussian Splatting and Neural Radiance Fields have become powerful tools for generating high-fidelity models of 3D objects and scenes. Their ability to produce both physically plausible and differentiable models of scenes are key ingredient needed to produce physically plausible adversarial attacks on DNNs. However, the adversarial machine learning community has yet to fully explore these capabilities, partly due to differing attack goals (e.g., misclassification, misdetection) and a wide range of possible scene manipulations used to achieve them (e.g., alter texture, mesh). This survey contributes the first framework that unifies diverse goals and tasks, facilitating easy comparison of existing work, identifying research gaps, and highlighting future directions - ranging from expanding attack goals and tasks to account for new modalities, state-of-the-art models, tools, and pipelines, to underscoring the importance of studying real-world threats in complex scenes.

cs.LG↗