Searcharxiv⌕ Search

arXiv subjects

Bin Wang

Publications and source records attributed to Bin Wang.

At least 253 records · Page 14Linked to original sources
arXiv2025

NextStep-1: Toward Autoregressive Image Generation with Continuous Tokens at Scale

Prevailing autoregressive (AR) models for text-to-image generation either rely on heavy, computationally-intensive diffusion models to process continuous image tokens, or employ vector quantization (VQ) to obtain discrete tokens with quantization loss. In this paper, we push the autoregressive paradigm forward with NextStep-1, a 14B autoregressive model paired with a 157M flow matching head, training on discrete text tokens and continuous image tokens with next-token prediction objectives. NextStep-1 achieves state-of-the-art performance for autoregressive models in text-to-image generation tasks, exhibiting strong capabilities in high-fidelity image synthesis. Furthermore, our method shows strong performance in image editing, highlighting the power and versatility of our unified approach. To facilitate open research, we will release our code and models to the community.

cs.CV↗
arXiv2025

Lumen: Consistent Video Relighting and Harmonious Background Replacement with Video Generative Models

Video relighting is a challenging yet valuable task, aiming to replace the background in videos while correspondingly adjusting the lighting in the foreground with harmonious blending. During translation, it is essential to preserve the original properties of the foreground, e.g., albedo, and propagate consistent relighting among temporal frames. In this paper, we propose Lumen, an end-to-end video relighting framework developed on large-scale video generative models, receiving flexible textual description for instructing the control of lighting and background. Considering the scarcity of high-qualified paired videos with the same foreground in various lighting conditions, we construct a large-scale dataset with a mixture of realistic and synthetic videos. For the synthetic domain, benefiting from the abundant 3D assets in the community, we leverage advanced 3D rendering engine to curate video pairs in diverse environments. For the realistic domain, we adapt a HDR-based lighting simulation to complement the lack of paired in-the-wild videos. Powered by the aforementioned dataset, we design a joint training curriculum to effectively unleash the strengths of each domain, i.e., the physical consistency in synthetic videos, and the generalized domain distribution in realistic videos. To implement this, we inject a domain-aware adapter into the model to decouple the learning of relighting and domain appearance distribution. We construct a comprehensive benchmark to evaluate Lumen together with existing methods, from the perspectives of foreground preservation and video consistency assessment. Experimental results demonstrate that Lumen effectively edit the input into cinematic relighted videos with consistent lighting and strict foreground preservation. Our project page: https://lumen-relight.github.io/

cs.CV↗
arXiv2025

A robust and compliant robotic assembly control strategy for batch precision assembly task with uncertain fit types and fit amounts

In some high-precision industrial applications, robots are deployed to perform precision assembly tasks on mass batches of manufactured pegs and holes. If the peg and hole are designed with transition fit, machining errors may lead to either a clearance or an interference fit for a specific pair of components, with uncertain fit amounts. This paper focuses on the robotic batch precision assembly task involving components with uncertain fit types and fit amounts, and proposes an efficient methodology to construct the robust and compliant assembly control strategy. Specifically, the batch precision assembly task is decomposed into multiple deterministic subtasks, and a force-vision fusion controller-driven reinforcement learning method and a multi-task reinforcement learning training method (FVFC-MTRL) are proposed to jointly learn multiple compliance control strategies for these subtasks. Subsequently, the multi-teacher policy distillation approach is designed to integrate multiple trained strategies into a unified student network, thereby establishing a robust control strategy. Real-world experiments demonstrate that the proposed method successfully constructs the robust control strategy for high-precision assembly task with different fit types and fit amounts. Moreover, the MTRL framework significantly improves training efficiency, and the final developed control strategy achieves superior force compliance and higher success rate compared with many existing methods.

cs.RO↗
arXiv2025

Winding Clearness for Differentiable Point Cloud Optimization

We propose to explore the properties of raw point clouds through the \emph{winding clearness}, a concept we first introduce for measuring the clarity of the interior/exterior relationships represented by the winding number field of the point cloud. In geometric modeling, the winding number is a powerful tool for distinguishing the interior and exterior of a given surface $\partial Ω$, and it has been previously used for point normal orientation and surface reconstruction. In this work, we introduce a novel approach to evaluate and optimize the quality of point clouds based on the winding clearness. We observe that point clouds with less noise generally exhibit better winding clearness. Accordingly, we propose an objective function that quantifies the error in winding clearness, solely utilizing the coordinates of the point clouds. Moreover, we demonstrate that the winding clearness error is differentiable and can serve as a loss function in point cloud processing. We present this observation from two aspects: 1) We update the coordinates of the points by back-propagating the loss function for individual point clouds, resulting in an overall improvement without involving a neural network. 2) We incorporate winding clearness as a geometric constraint in the diffusion-based 3D generative model and update the network parameters to generate point clouds with less noise. Experimental results demonstrate the effectiveness of optimizing the winding clearness in enhancing the point cloud quality. Notably, our method exhibits superior performance in handling noisy point clouds with thin structures, highlighting the benefits of the global perspective enabled by the winding number.

cs.GR↗
arXiv2025

Phase-engineered Non-degenerate Sliding Ferroelectricity Enables Tunable Photovoltaics in Monolayer Janus In2S2Se

Two-dimensional sliding ferroelectrics, with their enhanced efficiencies of charge separation and tunability, constitute promising platforms for next-generation photovoltaic devices. However, recent systems predominantly exhibit dual degenerate polarization states with weak intensity, hindering the optimal manipulations of photovoltaic effects through sliding ferroelectricity. Here, we address this limitation by introducing two strengthened and distinct non-degenerate sliding ferroelectric phases (WZ' and ZB') in Janus In2S2Se, which can be achieved by Se-to-S substitution in monolayer In2Se3. First-principles calculations validate the experimental synthesis of this structure and its capability for reversible phase transitions triggered by atomic layer sliding, and a series of superior photovoltaic performances are demonstrated in such unique Janus In2S2Se, accompanied by a detailed analysis of how non-degenerate sliding ferroelectricity modulates distinct photovoltaic characteristics. The WZ' to ZB' transition can increase the carrier mobility and moderate the band gap while inducing an indirect-to-direct transition, yielding a marked red-shift and enhancement of the photocurrent peak in the infrared spectrum. Conversely, the WZ' phase, benefiting from enhanced polarization, delivers superior photoelectric conversion efficiency in the visible light region. This work establishes a phase-engineered framework of how non-degenerate sliding ferroelectricity orchestrates distinct photovoltaic behaviors, and the intrinsic physical correlations may offer novel perspectives for designing and regulating innovative photovoltaic devices.

cond-mat.mtrl-sci↗
arXiv2025

Chimera: Improving Generalist Model with Domain-Specific Experts

Recent advancements in Large Multi-modal Models (LMMs) underscore the importance of scaling by increasing image-text paired data, achieving impressive performance on general tasks. Despite their effectiveness in broad applications, generalist models are primarily trained on web-scale datasets dominated by natural images, resulting in the sacrifice of specialized capabilities for domain-specific tasks that require extensive domain prior knowledge. Moreover, directly integrating expert models tailored for specific domains is challenging due to the representational gap and imbalanced optimization between the generalist model and experts. To address these challenges, we introduce Chimera, a scalable and low-cost multi-modal pipeline designed to boost the ability of existing LMMs with domain-specific experts. Specifically, we design a progressive training strategy to integrate features from expert models into the input of a generalist LMM. To address the imbalanced optimization caused by the well-aligned general visual encoder, we introduce a novel Generalist-Specialist Collaboration Masking (GSCM) mechanism. This results in a versatile model that excels across the chart, table, math, and document domains, achieving state-of-the-art performance on multi-modal reasoning and visual content extraction tasks, both of which are challenging tasks for assessing existing LMMs.

cs.CV↗
arXiv2025

ToolACE: Winning the Points of LLM Function Calling

Function calling significantly extends the application boundary of large language models, where high-quality and diverse training data is critical for unlocking this capability. However, real function-calling data is quite challenging to collect and annotate, while synthetic data generated by existing pipelines tends to lack coverage and accuracy. In this paper, we present ToolACE, an automatic agentic pipeline designed to generate accurate, complex, and diverse tool-learning data. ToolACE leverages a novel self-evolution synthesis process to curate a comprehensive API pool of 26,507 diverse APIs. Dialogs are further generated through the interplay among multiple agents, guided by a formalized thinking process. To ensure data accuracy, we implement a dual-layer verification system combining rule-based and model-based checks. We demonstrate that models trained on our synthesized data, even with only 8B parameters, achieve state-of-the-art performance on the Berkeley Function-Calling Leaderboard, rivaling the latest GPT-4 models. Our model and a subset of the data are publicly available at https://huggingface.co/Team-ACE.

cs.LG↗
arXiv2025

Max-Min Beamforming for Large-Scale Cell-Free Massive MIMO: A Randomized ADMM Algorithm

We consider the problem of max-min beamforming (MMB) for cell-free massive multi-input multi-output (MIMO) systems, where the objective is to maximize the minimum achievable rate among all users. Existing MMB methods are mainly based on deterministic optimization methods, which are computationally inefficient when the problem size grows large. To address this issue, we, in this paper, propose a randomized alternating direction method of multiplier (ADMM) algorithm for large-scale MMB problems. We first propose a novel formulation that transforms the highly challenging feasibility-checking problem into a linearly constrained optimization problem. An efficient randomized ADMM is then developed for solving the linearly constrained problem. Unlike standard ADMM, randomized ADMM only needs to solve a small number of subproblems at each iteration to ensure convergence, thus achieving a substantial complexity reduction. Our theoretical analysis reveals that the proposed algorithm exhibits an O(1/\bar{t}) convergence rate (\bar{t} represents the number of iterations), which is on the same order as its deterministic counterpart. Numerical results show that the proposed algorithm offers a significant complexity advantage over existing methods in solving the MMB problem.

eess.SP↗
arXiv2025

Step-3 is Large yet Affordable: Model-system Co-design for Cost-effective Decoding

Large language models (LLMs) face low hardware efficiency during decoding, especially for long-context reasoning tasks. This paper introduces Step-3, a 321B-parameter VLM with hardware-aware model-system co-design optimized for minimizing decoding costs. Step-3 innovates in two key dimensions: (1) A novel Multi-Matrix Factorization Attention (MFA) mechanism that significantly reduces both KV cache size and computation while maintaining high attention expressiveness, and (2) Attention-FFN Disaggregation (AFD), a distributed inference system that decouples attention and Feed-Forward Network (FFN) layers into specialized subsystems. This co-design achieves unprecedented cost efficiency: Step-3 significantly reduces theoretical decoding costs compared with models like DeepSeek-V3 and Qwen3 MoE 235B, with the gains widening at longer context. Step-3 achieves low cost while activating 38B parameters per token (more than DeepSeek-V3 and Qwen3 MoE 235B), demonstrating that hardware-aligned attention arithmetic intensity, MoE sparsity, and AFD are critical to cost-effectiveness. We perform a head-to-head comparison with DeepSeek-V3 in its favorable scenarios. Our implementation on Hopper GPUs achieves a decoding throughput of up to 4,039 tokens per second per GPU under 50ms TPOT SLA (4K context, FP8, no MTP). It is higher than DeepSeek-V3's 2,324 in the same setup and sets a new Pareto frontier for LLM decoding.

cs.LG↗
arXiv2025

Impact of Medium and Heavy-Duty Electric Vehicle Electrification on Distribution System Stability

Medium and heavy-duty (MHD) commercial vehicles contribute significantly to carbon emissions, accounting for 21\% of the total emissions in the transportation sector. To curb this, U.S. government is increasingly focusing on achieving 100\% fleet electrification over the next decade. However, the integration of megawatt-scale charging stations designed for MHD vehicles poses challenges to the stability of secondary distribution systems. This study investigates the impact of megawatt-scale charging station loads on a benchmark IEEE 33-bus distribution system using real data from the HEVI-LOAD software for MHD electrification planning developed by Lawrence Berkeley National Laboratory (LBNL). The results reveal significant violations of per-unit (p.u.) voltage values at various nodes of the distribution system, indicating that substantial upgrades to the distribution infrastructure will be necessary to accommodate the projected MHDEV charging loads and meet electrification targets.

nlin.AO↗
arXiv2025

Jamming-Resistant AAV Communications: A Multichannel-Aided Approach

Jamming cancellation is essential to reliable unmanned autonomous vehicle (AAV) communications in the presence of malicious jammers. In this paper, we develop a practical multichannel-aided jamming cancellation method to realize secure AAV communications. The proposed method is capable of simultaneously achieving timing/frequency synchronization as well as jamming cancellation. More importantly, our method does not need the signal's/jammer's channel state information. It only utilizes the knowledge of the legitimate sender's preamble sequence that is available in existing communication protocols. We also analyze the length of the preamble sequence required for successful synchronization and signal recovery. Experimental results on the built hardware platform show that, with a two-antenna receiver, the proposed method can successfully decode the signal of interest even when the jamming signal is $40$dB stronger than the communication signal.

eess.SP↗
arXiv2025

MiroMind-M1: An Open-Source Advancement in Mathematical Reasoning via Context-Aware Multi-Stage Policy Optimization

Large language models have recently evolved from fluent text generation to advanced reasoning across diverse domains, giving rise to reasoning language models. Among these domains, mathematical reasoning serves as a representative benchmark as it requires precise multi-step logic and abstract reasoning, which can be generalized to other tasks. While closed-source RLMs such as GPT-o3 demonstrate impressive reasoning capabilities, their proprietary nature limits transparency and reproducibility. Although many open-source projects aim to close this gap, most of them lack sufficient openness by omitting critical resources such as datasets and detailed training configurations, which hinders reproducibility. To contribute toward greater transparency in RLM development, we introduce the MiroMind-M1 series, a set of fully open-source RLMs built on the Qwen-2.5 backbone that match or exceed the performance of existing open-source RLMs. Specifically, our models are trained in two stages: SFT on a carefully curated corpus of 719K math-reasoning problems with verified CoT trajectories, followed by RLVR on 62K challenging and verifiable problems. To enhance the robustness and efficiency of the RLVR process, we introduce Context-Aware Multi-Stage Policy Optimization, an algorithm that integrates length-progressive training with an adaptive repetition penalty to encourage context-aware RL training. Our model achieves state-of-the-art or competitive performance and superior token efficiency among Qwen-2.5-based open-source 7B and 32B models on the AIME24, AIME25, and MATH benchmarks. To facilitate reproducibility, we release the complete stack: models (MiroMind-M1-SFT-7B, MiroMind-M1-RL-7B, MiroMind-M1-RL-32B); datasets (MiroMind-M1-SFT-719K, MiroMind-M1-RL-62K); and all training and evaluation configurations. We hope these resources will support further research and foster community advancement.

cs.CL↗
arXiv2025

Finite Groups of Symplectic Automorphisms of Supersingular K3 surfaces in Odd Characteristics

In 2009, Dolgachev-Keum showed that finite groups of tame symplectic automorphisms of K3 surfaces in positive characteristics are subgroups of the Mathieu group of degree 23. In this paper, we utilize lattice-theoretic methods to investigate symplectic actions of finite groups G on K3 surfaces in odd characteristics. For supersingular K3 surfaces with Artin invariants at least two, we develop a new machinery called p-root pairs to constrain possible symplectic finite group actions (without the assumption of tameness). The concept of p-root pair is closely related to root systems and Weyl groups. In particular, we provide alternative proof for many results by Dolgachev-Keum and give an upper bound for the exponent of p in |G|.

math.AG↗
arXiv2025

Function-to-Style Guidance of LLMs for Code Translation

Large language models (LLMs) have made significant strides in code translation tasks. However, ensuring both the correctness and readability of translated code remains a challenge, limiting their effective adoption in real-world software development. In this work, we propose F2STrans, a function-to-style guiding paradigm designed to progressively improve the performance of LLMs in code translation. Our approach comprises two key stages: (1) Functional learning, which optimizes translation correctness using high-quality source-target code pairs mined from online programming platforms, and (2) Style learning, which improves translation readability by incorporating both positive and negative style examples. Additionally, we introduce a novel code translation benchmark that includes up-to-date source code, extensive test cases, and manually annotated ground-truth translations, enabling comprehensive functional and stylistic evaluations. Experiments on both our new benchmark and existing datasets demonstrate that our approach significantly improves code translation performance. Notably, our approach enables Qwen-1.5B to outperform prompt-enhanced Qwen-32B and GPT-4 on average across 20 diverse code translation scenarios.

cs.AI↗
arXiv2025

Gaussian Fluids: A Grid-Free Fluid Solver based on Gaussian Spatial Representation

We present a grid-free fluid solver featuring a novel Gaussian representation. Drawing inspiration from the expressive capabilities of 3D Gaussian Splatting in multi-view image reconstruction, we model the continuous flow velocity as a weighted sum of multiple Gaussian functions. This representation is continuously differentiable, which enables us to derive spatial differentials directly and solve the time-dependent PDE via a custom first-order optimization tailored to fluid dynamics. Compared to traditional discretizations, which typically adopt Eulerian, Lagrangian, or hybrid perspectives, our approach is inherently memory-efficient and spatially adaptive, enabling it to preserve fine-scale structures and vortices with high fidelity. While these advantages are also sought by implicit neural representations, GSR offers enhanced robustness, accuracy, and generality across diverse fluid phenomena, with improved computational efficiency during temporal evolution. Though our first-order solver does not yet match the speed of fluid solvers using explicit representations, its continuous nature substantially reduces spatial discretization error and opens a new avenue for high-fidelity simulation. We evaluate the proposed solver across a broad range of 2D and 3D fluid phenomena, demonstrating its ability to preserve intricate vortex dynamics, accurately capture boundary-induced effects such as Kármán vortex streets, and remain robust across long time horizons - all without additional parameter tuning. Our results suggest that GSR offers a compelling direction for future research in fluid simulation.

cs.GR↗
arXiv2025

NTU Speechlab LLM-Based Multilingual ASR System for Interspeech MLC-SLM Challenge 2025

This report details the NTU Speechlab system developed for the Interspeech 2025 Multilingual Conversational Speech and Language Model (MLC-SLM) Challenge (Task I), where we achieved 5th place. We present comprehensive analyses of our multilingual automatic speech recognition system, highlighting key advancements in model architecture, data selection, and training strategies. In particular, language-specific prompts and model averaging techniques were instrumental in boosting system performance across diverse languages. Compared to the initial baseline system, our final model reduced the average Mix Error Rate from 20.2% to 10.6%, representing an absolute improvement of 9.6% (a relative improvement of 48%) on the evaluation set. Our results demonstrate the effectiveness of our approach and offer practical insights for future Speech Large Language Models.

cs.CL↗
arXiv2025

Derivative-Free Optimization-Empowered Wireless Channel Reconfiguration for 6G

Reconfigurable antennas, including reconfigurable intelligent surface (RIS), movable antenna (MA), fluid antenna (FA), and other advanced antenna techniques, have been studied extensively in the context of reshaping wireless propagation environments for 6G and beyond wireless communications. Nevertheless, how to reconfigure/optimize the real-time controllable coefficients to achieve a favorable end-to-end wireless channel remains a substantial challenge, as it usually requires accurate modeling of the complex interaction between the reconfigurable devices and the electromagnetic waves, as well as knowledge of implicit channel propagation parameters. In this paper, we introduce a derivative-free optimization (a.k.a., zeroth-order (ZO) optimization) technique to directly optimize reconfigurable coefficients to shape the wireless end-to-end channel, without the need of channel modeling and estimation of the implicit environmental propagation parameters. We present the fundamental principles of ZO optimization and discuss its potential advantages in wireless channel reconfiguration. Two case studies for RIS and movable antenna-enabled single-input single-output (SISO) systems are provided to show the superiority of ZO-based methods as compared to state-of-the-art techniques. Finally, we outline promising future research directions and offer concluding insights on derivative-free optimization for reconfigurable antenna technologies.

eess.SP↗