arXiv subjects

Bowen Peng

Publications and source records attributed to Bowen Peng.

At least 19 recordsLinked to original sources

arXiv2026

SARATR-X-v2: Scale-Aware Structural Pre-Training for SAR Foundation Models

Masked image modeling has become a dominant paradigm for SAR pre-training, yet the design of the reconstruction target remains fundamentally unsettled. This article argues that a SAR pre-training target should satisfy two conditions to produce transferable representations: (i) physics-grounded stability, i.e., approximate invariance of the target operator to multiplicative speckle inherent in coherent imaging; and (ii) semantic scale compatibility, i.e., coverage of the heterogeneous spatial scales that downstream tasks demand. These two conditions are individually achievable but jointly difficult: physics-grounded stability favors fixed operators, while semantic scale compatibility favors data-driven composition. To this end, SARATR-X-v2 reconciles both within a single design. The target is constructed through fixed structural extractors spanning six receptive fields, from blind-spot local aggregation to directional log-ratio region contrast, and fused via learnable weights into one unified supervision signal for masked reconstruction. On twelve SAR benchmarks across classification, detection, and segmentation, SARATR-X-v2 achieves state-of-the-art transfer performance. Under synthetic speckle variation, the proposed target reduces perturbation drift in the learned representation by nearly two orders of magnitude relative to pixel-space supervision. Taken together, these results establish physics-grounded stability and semantic scale compatibility as a principled framework for pre-training target design under coherent imaging, and suggest that effective SAR pre-training is not about reconstructing more signal, but about reconstructing the right structural target.

cs.CV↗

arXiv2026

Multi-Token Residual Prediction

Diffusion Language Models (DLMs) generate text by iteratively denoising masked token sequences, offering a tradeoff between parallelism and quality compared to autoregressive models. In current practice, the number of tokens decoded per step is controlled by a confidence threshold, and quality degrades monotonically as more tokens are denoised per step. We introduce Multi-token Residual Prediction (MRP), a lightweight module that enables dependency-aware multi-token denoising within a single backbone forward pass. MRP exploits a key property of the denoising process: the logit distributions at adjacent denoising steps are remarkably similar. Rather than running the backbone a second time to obtain the next-step logits, MRP predicts the residual between steps from the backbone's hidden states, effectively denoising more tokens per backbone forward at a fraction of the cost. We apply MRP across the two operating regimes of DLM decoding. In the high-quality-low-throughput static denoising regime, MRP serves as a drafter for speculative decoding: its proposals are verified against the backbone, yielding lossless acceleration of up to 1.4x in SGLang. In the low-quality-high-throughput dynamic denoising regime, MRP instead drives a remasking scheme that revokes over-eager reveals, recovering most of the accuracy lost to aggressive low-threshold decoding and improving accuracy by up to 22.6 points on code generation task HumanEval and 17.7 points on reasoning task GSM8K.

cs.LG↗

arXiv2026

RePO-VLA: Recovery-Driven Policy Optimization for Vision-Language-Action Models

Vision-Language-Action (VLA) models remain brittle in long-horizon, contact-rich manipulation because success-only imitation provides little supervision for execution drift, while failed rollouts are often discarded. We introduce RePO-VLA, a recovery-driven policy optimization framework that assigns distinct roles to success, recovery, and failure trajectories. RePO-VLA first applies Recovery-Aware Initialization (RAI), slicing recovery segments and resetting history so corrective actions depend on the current adverse state rather than the preceding failure. It then learns a Progress-Aware Semantic Value Function (PAS-VF), aligning spatiotemporal trajectory features with instructions and successful references. The resulting labels salvage useful failure prefixes via reliability decay, while low-value labels mark drift and terminal breakdowns, teaching differences among nominal, failed, and corrective actions. The data engine turns adverse states into planner-generated or human-collected corrective rollouts, teaching recovery to the success manifold. Value-Conditioned Refinement (VCR) trains the policy to prefer high-progress actions. At deployment, a fixed high value ($v=1.0$) biases actions toward the learned success manifold without online failure detectors or heuristic retries. We introduce FRBench, with standardized error injection and recovery-focused evaluation. Across simulated and real-world bimanual tasks, RePO-VLA improves robustness, raising adversarial success from 20% to 75% on average and up to 80% in scaled real-world trials.

cs.RO↗

arXiv2026

Efficient Pre-Training with Token Superposition

Pre-training of Large Language Models is often prohibitively expensive and inefficient at scale, requiring complex and invasive modifications in order to achieve high data throughput. In this work, we present Token-Superposition Training (TST), a simple drop-in method that significantly improves the data throughput per FLOPs during pre-training without modifying the parallelism, optimizer, tokenizer, data, or model architecture. TST is done in two phases: (i) A highly efficient superposition phase where we combine many contiguous tokens into one bag and train using a multi-hot cross-entropy (MCE) objective, and (ii) a recovery phase where we revert back to standard training. We extensively evaluate TST on the scale of 270M and 600M parameters and validate on 3B and a 10B A1B mixture of experts model, demonstrating that it is highly robust in different settings. Ultimately, TST consistently outperforms baseline loss and downstream evaluations, and under equal-loss settings, TST yields up to a 2.5x reduction in total pre-training time at the 10B A1B scale.

cs.CL↗

arXiv2026

Long Context Pre-Training with Lighthouse Attention

Training causal transformers at extreme sequence lengths is bottlenecked by the quadratic time and memory of scaled dot-product attention (SDPA). In this work, we propose Lighthouse Attention, a training-only symmetrical selection-based hierarchical attention algorithm that wraps around ordinary SDPA and can be easily removed towards the end of the training. Our hierarchical selection is also gradient-free, which exempts us from dealing with a complicated and potentially inefficient backward pass kernel. Our contribution is three-fold: (i) A subquadratic hierarchical pre- and post-processing step that does adaptive compression and decompression of the sequence. (ii) A symmetrical compression strategy that pools queries, keys and values at the same time, while preserving left-to-right causality, which greatly improves parallelism. (iii) A two stage training approach which we pre-train for the majority of the time with Lighthouse Attention and recover a full attention model at the end with a short training. We run preliminary small scale LLM pre-training experiments that show the effectiveness of our method compared to full attention training with all other settings matched, where we achieve a faster total training time and lower final loss after the recovery phase. Full code is available at: https://github.com/ighoshsubho/lighthouse-attention

cs.CL↗

arXiv2026

Decoupling the Benefits of Subword Tokenization for Language Model Training via Byte-level Simulation

Subword tokenization is an essential part of modern large language models (LLMs), yet its specific contributions to training efficiency and model performance remain poorly understood. In this work, we decouple the effects of subword tokenization by isolating them within a controlled byte-level pretraining pipeline. We formulate and test hypotheses across various dimensions, including sample throughput, vocabulary scaling, and the linguistic prior of subword boundaries. By simulating these effects in a byte-level setting, we refine our understanding of why subword models outperform raw byte models and offer insights to improve the pretraining of future byte-level and subword models. Specifically, our experiments highlight the critical role of increased training throughput and the integration of subword boundaries as either explicit priors or inductive biases.

cs.CL↗

arXiv2026

Better with Less: Tackling Heterogeneous Multi-Modal Image Joint Pretraining via Conditioned and Degraded Masked Autoencoder

Learning robust representations across extremely heterogeneous modalities remains a fundamental challenge in multi-modal vision. As a critical and profound instantiation of this challenge, high-resolution (HR) joint optical and synthetic aperture radar (SAR) pretraining seeks modality synergy to mutually enhance single-source representations; its potential is severely hindered by the Heterogeneity-Resolution Paradox: finer spatial scales drastically amplify the physical divergence between complex radar geometries and non-homologous optical textures. Consequently, migrating medium-resolution-oriented rigid alignment paradigms to HR scenarios triggers either severe feature suppression to force equivalence, or feature contamination driven by extreme epistemic uncertainty. Both extremes inevitably culminate in profound representation degradation and negative transfer. To overcome this bottleneck, we propose CoDe-MAE, pioneering a \textit{better synergy with less alignment} philosophy. First, Optical-anchored Knowledge Distillation (OKD) implicitly regularizes SAR's speckle noise by mapping it into a pure semantic manifold. Building on this, Conditioned Contrastive Learning (CCL) utilizes a gradient buffering mechanism to align shared consensus while safely preserving divergent physical signatures. Concurrently, Cross-Modal Degraded Reconstruction (CDR) deliberately strips non-homologous spectral pseudo-features, truncating the inherently ill-posed mapping to capture true structural invariants. Extensive analyses validate our theoretical claims. Pretrained on 1M samples, CoDe-MAE demonstrates remarkable data efficiency, successfully preventing representation degradation and establishing new state-of-the-art performance across diverse single- and bi-modal downstream tasks, substantially outperforming foundation models scaled on vastly larger datasets.

cs.CV↗

arXiv2026

Distilling Token-Trained Models into Byte-Level Models

Byte Language Models (BLMs) have emerged as a promising direction for scaling language models beyond tokenization. However, existing BLMs typically require training from scratch on trillions of bytes, making them prohibitively expensive. In this paper, we propose an efficient distillation recipe that converts existing token-trained LLMs into BLMs while retaining comparable capabilities. Our recipe follows a two-stage curriculum: (1) Progressive Knowledge Distillation, which aligns byte-level representations with the embeddings of the token-trained teacher model; and (2) Byte-Level Supervised Fine-Tuning, which enables end-to-end generation entirely in the byte space. We validate our approach across multiple model families, including Llama, Qwen, and OLMo, and demonstrate that the distilled BLMs retain most of the teacher models' performance using only approximately 125B bytes.

cs.CL↗

arXiv2025

Fifty Years of SAR Automatic Target Recognition: The Road Forward

Synthetic Aperture Radar (SAR) imaging is capable of observing objects in nearly all weather and illumination conditions, and has become an indispensable means of information acquisition for analysis and recognition of objects and scenes. SAR Automatic Target Recognition (SAR ATR) has been one of the most fundamental and challenging problems in remote sensing image analysis. Nowadays, the artificial intelligence (AI) technology, represented by large models and AI agents, has transformed the research paradigm, profoundly influenced various research fields, and continues to evolve at an unprecedented pace. However, the huge potential of AI for SAR image analysis remains locked. To unlock the potential of AI in SAR image understanding, the research community should rethink how to enable bidirectional empowerment between AI and SAR image understanding and strive to achieve substantial breakthroughs at critical bottlenecks. Given this period of remarkable evolution, this paper offers the first comprehensive review of SAR ATR, tracing its development and milestones over the past five decades and providing the research community with a clear roadmap. This survey includes approximately 260 research contributions, covering critical aspects of SAR ATR: pivotal challenges, important datasets, the merits and limitations of representative methods, evaluation metrics, and state-of-the-art performance. Finally, we finish the survey by identifying promising directions for future research. Looking ahead, we call for significant attention on three fundamental pillars: the curation of high-quality large-scale datasets, the design of fair and comprehensive evaluation benchmarks, and the fostering of safe open-source ecosystems.

eess.IV↗

arXiv2025

Constraining fuzzy dark matter with the 21-cm power spectrum from Cosmic Dawn and Reionization

The 21-cm signals from Cosmic Dawn and the Epoch of Reionization contain valuable information on cosmological structure formation dominated by dark matter. Measurements of the 21-cm power spectrum can thus probe certain dark matter candidates. Here we investigate the impacts of fuzzy dark matter (FDM) on the 21-cm signals, taking into account both the linear matter power spectrum and the halo mass function (HMF) in FDM cosmologies. The full FDM dynamics are implemented in reionization simulations, along with a new ansatz on modulation of the FDM HMF by the linear overdensity. Not only does the suppression of FDM halos on small scales give rise to delay of the signature epochs during cosmic reionization, but these epochs are also shortened relative to the cold dark matter cosmology. In addition, we find that while the FDM effects on the 21-cm power spectrum are dominated by its linear dynamics early in Cosmic Dawn, a correct FDM HMF resulting from nonlinear wave dynamics must be considered when X-ray heating begins. We forecast the constraints on the FDM model parameters from upcoming 21-cm power spectrum measurements by SKA1-Low (central area). In FDM cosmologies with $m_\mathrm{FDM}=10^{-21}$ eV, SKA1-Low will be able to constrain the boson mass to within $\sim10$% at 2$\sigma$ confidence with a mock 1080-hour observation, if the ionizing efficiency is mass independent. However, our results show that realistic astrophysical processes are degenerate with the FDM effects, which shall severely loosen the constraints on the boson mass from 21-cm power spectrum data alone.

astro-ph.CO↗

arXiv2025

ATRNet-STAR: A Large Dataset and Benchmark Towards Remote Sensing Object Recognition in the Wild

The absence of publicly available, large-scale, high-quality datasets for Synthetic Aperture Radar Automatic Target Recognition (SAR ATR) has significantly hindered the application of rapidly advancing deep learning techniques, which hold huge potential to unlock new capabilities in this field. This is primarily because collecting large volumes of diverse target samples from SAR images is prohibitively expensive, largely due to privacy concerns, the characteristics of microwave radar imagery perception, and the need for specialized expertise in data annotation. Throughout the history of SAR ATR research, there have been only a number of small datasets, mainly including targets like ships, airplanes, buildings, etc. There is only one vehicle dataset MSTAR collected in the 1990s, which has been a valuable source for SAR ATR. To fill this gap, this paper introduces a large-scale, new dataset named ATRNet-STAR with 40 different vehicle categories collected under various realistic imaging conditions and scenes. It marks a substantial advancement in dataset scale and diversity, comprising over 190,000 well-annotated samples, 10 times larger than its predecessor, the famous MSTAR. Building such a large dataset is a challenging task, and the data collection scheme will be detailed. Secondly, we illustrate the value of ATRNet-STAR via extensively evaluating the performance of 15 representative methods with 7 different experimental settings on challenging classification and detection benchmarks derived from the dataset. Finally, based on our extensive experiments, we identify valuable insights for SAR ATR and discuss potential future research directions in this field. We hope that the scale, diversity, and benchmark of ATRNet-STAR can significantly facilitate the advancement of SAR ATR.

cs.CV↗

arXiv2024

DeMo: Decoupled Momentum Optimization

Scaling neural network training increasingly depends on synchronous data-parallelism, yet full-precision gradient all-reduce imposes a severe communication bottleneck. We propose Decoupled Momentum Optimization (DeMo), a drop-in replacement for any momentum-based optimizers that significantly reduces the communication bandwidth while maintaining convergence. DeMo (i) decouples local momentum updates, (ii) applies a fast orthonormal transform (e.g., DCT) followed by top-k sparsification, and (iii) reuses the momentum buffer as error feedback via momentum subtraction. This design reduces per-step communication by up to two orders of magnitude with minimal computational overhead. Experiments on 300M and 1B-parameter DeMo language models show DeMo transmits up to 85x less data per GPU than AdamW-DDP while achieving comparable loss and accuracy. DeMo is topology-agnostic and enables training across multi-datacenter or Ethernet-based setups. Code is available at https://github.com/bloc97/DeMo

cs.LG↗

arXiv2024

MaDiNet: Mamba Diffusion Network for SAR Target Detection

The fundamental challenge in SAR target detection lies in developing discriminative, efficient, and robust representations of target characteristics within intricate non-cooperative environments. However, accurate target detection is impeded by factors including the sparse distribution and discrete features of the targets, as well as complex background interference. In this study, we propose a \textbf{Ma}mba \textbf{Di}ffusion \textbf{Net}work (MaDiNet) for SAR target detection. Specifically, MaDiNet conceptualizes SAR target detection as the task of generating the position (center coordinates) and size (width and height) of the bounding boxes in the image space. Furthermore, we design a MambaSAR module to capture intricate spatial structural information of targets and enhance the capability of the model to differentiate between targets and complex backgrounds. The experimental results on extensive SAR target detection datasets achieve SOTA, proving the effectiveness of the proposed network. Code is available at \href{https://github.com/JoyeZLearning/MaDiNet}{https://github.com/JoyeZLearning/MaDiNet}.

eess.IV↗

arXiv2024

S$^4$ST: A Strong, Self-transferable, faSt, and Simple Scale Transformation for Transferable Targeted Attack

Transferable Targeted Attacks (TTAs) face significant challenges due to severe overfitting to surrogate models. Recent breakthroughs heavily rely on large-scale training data of victim models, while data-free solutions, \textit{i.e.}, image transformation-involved gradient optimization, often depend on black-box feedback for method design and tuning. These dependencies violate black-box transfer settings and compromise threat evaluation fairness. In this paper, we propose two blind estimation measures, self-alignment and self-transferability, to analyze per-transformation effectiveness and cross-transformation correlations under strict black-box constraints. Our findings challenge conventional assumptions: (1) Attacking simple scaling transformations uniquely enhances targeted transferability, outperforming other basic transformations and rivaling leading complex methods; (2) Geometric and color transformations exhibit high internal redundancy despite weak inter-category correlations. These insights drive the design and tuning of S$^4$ST (Strong, Self-transferable, faSt, Simple Scale Transformation), which integrates dimensionally consistent scaling, complementary low-redundancy transformations, and block-wise operations. Extensive evaluations across diverse architectures, training distributions, and tasks show that S$^{4}$ST achieves state-of-the-art effectiveness-efficiency balance without data dependency. We reveal that scaling's effectiveness stems from visual data's multi-scale nature and ubiquitous scale augmentation during training, rendering such augmentation a double-edged sword. Further validations on medical imaging and face verification confirm the framework's strong generalization.

cs.CR↗

arXiv2024

Enhancing Transferability of Targeted Adversarial Examples: A Self-Universal Perspective

Transfer-based targeted adversarial attacks against black-box deep neural networks (DNNs) have been proven to be significantly more challenging than untargeted ones. The impressive transferability of current SOTA, the generative methods, comes at the cost of requiring massive amounts of additional data and time-consuming training for each targeted label. This results in limited efficiency and flexibility, significantly hindering their deployment in practical applications. In this paper, we offer a self-universal perspective that unveils the great yet underexplored potential of input transformations in pursuing this goal. Specifically, transformations universalize gradient-based attacks with intrinsic but overlooked semantics inherent within individual images, exhibiting similar scalability and comparable results to time-consuming learning over massive additional data from diverse classes. We also contribute a surprising empirical insight that one of the most fundamental transformations, simple image scaling, is highly effective, scalable, sufficient, and necessary in enhancing targeted transferability. We further augment simple scaling with orthogonal transformations and block-wise applicability, resulting in the Simple, faSt, Self-universal yet Strong Scale Transformation (S$^4$ST) for self-universal TTA. On the ImageNet-Compatible benchmark dataset, our method achieves a 19.8% improvement in the average targeted transfer success rate against various challenging victim models over existing SOTA transformation methods while only consuming 36% time for attacking. It also outperforms resource-intensive attacks by a large margin in various challenging settings.

cs.CV↗

arXiv2024

Towards Assessing the Synthetic-to-Measured Adversarial Vulnerability of SAR ATR

Recently, there has been increasing concern about the vulnerability of deep neural network (DNN)-based synthetic aperture radar (SAR) automatic target recognition (ATR) to adversarial attacks, where a DNN could be easily deceived by clean input with imperceptible but aggressive perturbations. This paper studies the synthetic-to-measured (S2M) transfer setting, where an attacker generates adversarial perturbation based solely on synthetic data and transfers it against victim models trained with measured data. Compared with the current measured-to-measured (M2M) transfer setting, our approach does not need direct access to the victim model or the measured SAR data. We also propose the transferability estimation attack (TEA) to uncover the adversarial risks in this more challenging and practical scenario. The TEA makes full use of the limited similarity between the synthetic and measured data pairs for blind estimation and optimization of S2M transferability, leading to feasible surrogate model enhancement without mastering the victim model and data. Comprehensive evaluations based on the publicly available synthetic and measured paired labeled experiment (SAMPLE) dataset demonstrate that the TEA outperforms state-of-the-art methods and can significantly enhance various attack algorithms in computer vision and remote sensing applications. Codes and data are available at https://github.com/scenarri/S2M-TEA.

cs.CV↗

arXiv2023

ARBiBench: Benchmarking Adversarial Robustness of Binarized Neural Networks

Network binarization exhibits great potential for deployment on resource-constrained devices due to its low computational cost. Despite the critical importance, the security of binarized neural networks (BNNs) is rarely investigated. In this paper, we present ARBiBench, a comprehensive benchmark to evaluate the robustness of BNNs against adversarial perturbations on CIFAR-10 and ImageNet. We first evaluate the robustness of seven influential BNNs on various white-box and black-box attacks. The results reveal that 1) The adversarial robustness of BNNs exhibits a completely opposite performance on the two datasets under white-box attacks. 2) BNNs consistently exhibit better adversarial robustness under black-box attacks. 3) Different BNNs exhibit certain similarities in their robustness performance. Then, we conduct experiments to analyze the adversarial robustness of BNNs based on these insights. Our research contributes to inspiring future research on enhancing the robustness of BNNs and advancing their application in real-world scenarios.

cs.CV↗

arXiv2023

YaRN: Efficient Context Window Extension of Large Language Models

Rotary Position Embeddings (RoPE) have been shown to effectively encode positional information in transformer-based language models. However, these models fail to generalize past the sequence length they were trained on. We present YaRN (Yet another RoPE extensioN method), a compute-efficient method to extend the context window of such models, requiring 10x less tokens and 2.5x less training steps than previous methods. Using YaRN, we show that LLaMA models can effectively utilize and extrapolate to context lengths much longer than their original pre-training would allow, while also surpassing previous the state-of-the-art at context window extension. In addition, we demonstrate that YaRN exhibits the capability to extrapolate beyond the limited context of a fine-tuning dataset. Code is available at https://github.com/jquesnelle/yarn

cs.CL↗