arXiv subjects

Bo Wang

Publications and source records attributed to Bo Wang.

At least 181 records · Page 10Linked to original sources

arXiv2026

A 96pJ/Frame/Pixel and 61pJ/Event Anti-UAV System with Hybrid Object Tracking Modes

We present an energy-efficient anti-UAV system that integrates frame-based and event-driven object tracking to enable reliable detection of small and fast-moving drones. The system reconstructs binary event frames using run-length encoding, generates region proposals, and adaptively switches between frame mode and event mode based on object size and velocity. A Fast Object Tracking Unit improves robustness for high-speed targets through adaptive thresholding and trajectory-based classification. The neural processing unit supports both grayscale-patch and trajectory inference with a custom instruction set and a zero-skipping MAC architecture, reducing redundant neural computations by more than 97 percent. Implemented in 40 nm CMOS technology, the 2 mm^2 chip achieves 96 pJ per frame per pixel and 61 pJ per event at 0.8 V, and reaches 98.2 percent recognition accuracy on public UAV datasets across 50 to 400 m ranges and 5 to 80 pixels per second speeds. The results demonstrate state-of-the-art end-to-end energy efficiency for anti-UAV systems.

cs.CV↗

arXiv2026

EAG3R: Event-Augmented 3D Geometry Estimation for Dynamic and Extreme-Lighting Scenes

Robust 3D geometry estimation from videos is critical for applications such as autonomous navigation, SLAM, and 3D scene reconstruction. Recent methods like DUSt3R demonstrate that regressing dense pointmaps from image pairs enables accurate and efficient pose-free reconstruction. However, existing RGB-only approaches struggle under real-world conditions involving dynamic objects and extreme illumination, due to the inherent limitations of conventional cameras. In this paper, we propose EAG3R, a novel geometry estimation framework that augments pointmap-based reconstruction with asynchronous event streams. Built upon the MonST3R backbone, EAG3R introduces two key innovations: (1) a retinex-inspired image enhancement module and a lightweight event adapter with SNR-aware fusion mechanism that adaptively combines RGB and event features based on local reliability; and (2) a novel event-based photometric consistency loss that reinforces spatiotemporal coherence during global optimization. Our method enables robust geometry estimation in challenging dynamic low-light scenes without requiring retraining on night-time data. Extensive experiments demonstrate that EAG3R significantly outperforms state-of-the-art RGB-only baselines across monocular depth estimation, camera pose tracking, and dynamic reconstruction tasks.

cs.CV↗

arXiv2026

SetPO: Set-Level Policy Optimization for Diversity-Preserving LLM Reasoning

Reinforcement learning with verifiable rewards has shown notable effectiveness in enhancing large language models (LLMs) reasoning performance, especially in mathematics tasks. However, such improvements often come with reduced outcome diversity, where the model concentrates probability mass on a narrow set of solutions. Motivated by diminishing-returns principles, we introduce a set level diversity objective defined over sampled trajectories using kernelized similarity. Our approach derives a leave-one-out marginal contribution for each sampled trajectory and integrates this objective as a plug-in advantage shaping term for policy optimization. We further investigate the contribution of a single trajectory to language model diversity within a distribution perturbation framework. This analysis theoretically confirms a monotonicity property, proving that rarer trajectories yield consistently higher marginal contributions to the global diversity. Extensive experiments across a range of model scales demonstrate the effectiveness of our proposed algorithm, consistently outperforming strong baselines in both Pass@1 and Pass@K across various benchmarks.

cs.AI↗

arXiv2026

SONIC: Segmented Optimized Nexus for Information Compression in Key-Value Caching

The linear growth of Key-Value (KV) cache remains a bottleneck for multi-turn LLM deployment. Existing KV cache compression methods often fail to account for the structural properties of multi-turn dialogues, relying on heuristic eviction that risks losing critical context. We propose \textbf{SONIC}, a learning-based framework that compresses historical segments into compact and semantically rich \textbf{Nexus} tokens. By integrating dynamic budget training, SONIC allows flexible adaptation to varying memory constraints without retraining. Experiments show that at compression ratios of 80\% and 50\%, SONIC consistently outperforms baselines such as H2O and StreamingLLM on four diverse multi-turn benchmarks. Specifically, on the widely used MTBench101 benchmark, SONIC achieves an average score improvement of 35.55\% over state-of-the-art baselines, validating its effectiveness in sustaining coherent multi-turn dialogues. Furthermore, SONIC enhances deployment efficiency, accelerating the overall inference process by 50.1\% compared to full-context generation.

cs.CL↗

arXiv2026

Explicit Multi-head Attention for Inter-head Interaction in Large Language Models

In large language models built upon the Transformer architecture, recent studies have shown that inter-head interaction can enhance attention performance. Motivated by this, we propose Multi-head Explicit Attention (MEA), a simple yet effective attention variant that explicitly models cross-head interaction. MEA consists of two key components: a Head-level Linear Composition (HLC) module that separately applies learnable linear combinations to the key and value vectors across heads, thereby enabling rich inter-head communication; and a head-level Group Normalization layer that aligns the statistical properties of the recombined heads. MEA shows strong robustness in pretraining, which allows the use of larger learning rates that lead to faster convergence, ultimately resulting in lower validation loss and improved performance across a range of tasks. Furthermore, we explore the parameter efficiency of MEA by reducing the number of attention heads and leveraging HLC to reconstruct them using low-rank "virtual heads". This enables a practical key-value cache compression strategy that reduces KV-cache memory usage by 50% with negligible performance loss on knowledge-intensive and scientific reasoning tasks, and only a 3.59% accuracy drop for Olympiad-level mathematical benchmarks.

cs.LG↗

arXiv2026

TL-GRPO: Turn-Level RL for Reasoning-Guided Iterative Optimization

Large language models have demonstrated strong reasoning capabilities in complex tasks through tool integration, which is typically framed as a Markov Decision Process and optimized with trajectory-level RL algorithms such as GRPO. However, a common class of reasoning tasks, iterative optimization, presents distinct challenges: the agent interacts with the same underlying environment state across turns, and the value of a trajectory is determined by the best turn-level reward rather than cumulative returns. Existing GRPO-based methods cannot perform fine-grained, turn-level optimization in such settings, while black-box optimization methods discard prior knowledge and reasoning capabilities. To address this gap, we propose Turn-Level GRPO (TL-GRPO), a lightweight RL algorithm that performs turn-level group sampling for fine-grained optimization. We evaluate TL-GRPO on analog circuit sizing (ACS), a challenging scientific optimization task requiring multiple simulations and domain expertise. Results show that TL-GRPO outperforms standard GRPO and Bayesian optimization methods across various specifications. Furthermore, our 30B model trained with TL-GRPO achieves state-of-the-art performance on ACS tasks under same simulation budget, demonstrating both strong generalization and practical utility.

cs.CL↗

arXiv2026

Bootstrap AdS Veneziano Amplitude with Arbitrary Kaluza-Klein Modes

We present a derivation of the first curvature correction to the AdS Veneziano amplitude for arbitrary Kaluza-Klein (KK) modes, using a bootstrap approach based on the world-sheet representation and AdS$\times$S formalism. Our results establish a universal formula for the first order curvature correction without considering any low-lying KK configurations. We give new predictions for Wilson coefficients in the low-energy expansion. In the high-energy regime, the amplitude exhibits a universal exponent independent of the external KK charges, providing a coherent picture of AdS stringy amplitudes in different backgrounds.

hep-th↗

arXiv2026

A Comprehensive Study on Large Language Models for Mutation Testing

Large Language Models (LLMs) have recently been used to generate mutants in both research work and in industrial practice. However, there has been no comprehensive empirical study of their performance for this increasingly important LLM-based Software Engineering application. To address this, we conduct a comprehensive empirical study evaluating BugFarm and LLMorpheus (the two state-of-the-art LLM-based approaches), alongside seven LLMs using our newly designed prompt, including both leading open- and closed-source models, on 851 real bugs from two Java real-world bug benchmarks. Our results reveal that, compared to existing rule-based approaches, LLMs generate more diverse mutants, that are behaviorally closer to real bugs and, most importantly, with 111.29% higher fault detection. That is, 87.98% (for LLMs) vs. 41.64% (for rule-based); an increase of 46.34 percentage points. Nevertheless, our results also reveal that these impressive results for improved effectiveness come at a cost: the LLM-generated mutants have worse non-compilability, duplication, and equivalent mutant rates by 26.60, 10.14, and 3.51 percentage points, respectively. These findings are immediately actionable for both research and practice. They allow practitioners to have greater confidence in deploying LLM-based mutation, while researchers now have a baseline for the state-of-the-art, with which they can research techniques to further improve effectiveness and reduce cost.

cs.SE↗

arXiv2026

ABC-Bench: Benchmarking Agentic Backend Coding in Real-World Development

The evolution of Large Language Models (LLMs) into autonomous agents has expanded the scope of AI coding from localized code generation to complex, repository-level, and execution-driven problem solving. However, current benchmarks predominantly evaluate code logic in static contexts, neglecting the dynamic, full-process requirements of real-world engineering, particularly in backend development which demands rigorous environment configuration and service deployment. To address this gap, we introduce ABC-Bench, a benchmark explicitly designed to evaluate agentic backend coding within a realistic, executable workflow. Using a scalable automated pipeline, we curated 224 practical tasks spanning 8 languages and 19 frameworks from open-source repositories. Distinct from previous evaluations, ABC-Bench require the agents to manage the entire development lifecycle from repository exploration to instantiating containerized services and pass the external end-to-end API tests. Our extensive evaluation reveals that even state-of-the-art models struggle to deliver reliable performance on these holistic tasks, highlighting a substantial disparity between current model capabilities and the demands of practical backend engineering. Our code is available at https://github.com/OpenMOSS/ABC-Bench.

cs.SE↗

arXiv2026

From Performance to Practice: Knowledge-Distilled Segmentator for On-Premises Clinical Workflows

Deploying medical image segmentation models in routine clinical workflows is often constrained by on-premises infrastructure, where computational resources are fixed and cloud-based inference may be restricted by governance and security policies. While high-capacity models achieve strong segmentation accuracy, their computational demands hinder practical deployment and long-term maintainability in hospital environments. We present a deployment-oriented framework that leverages knowledge distillation to translate a high-performing segmentation model into a scalable family of compact student models, without modifying the inference pipeline. The proposed approach preserves architectural compatibility with existing clinical systems while enabling systematic capacity reduction. The framework is evaluated on a multi-site brain MRI dataset comprising 1,104 3D volumes, with independent testing on 101 curated cases, and is further examined on abdominal CT to assess cross-modality generalizability. Under aggressive parameter reduction (94%), the distilled student model preserves nearly all of the teacher's segmentation accuracy (98.7%), while achieving substantial efficiency gains, including up to a 67% reduction in CPU inference latency without additional deployment overhead. These results demonstrate that knowledge distillation provides a practical and reliable pathway for converting research-grade segmentation models into maintainable, deployment-ready components for on-premises clinical workflows in real-world health systems.

cs.CV↗

arXiv2026

Deconstructing Pre-training: Knowledge Attribution Analysis in MoE and Dense Models

Mixture-of-Experts (MoE) architectures decouple model capacity from per-token computation, enabling scaling beyond the computational limits imposed by dense scaling laws. Yet how MoE architectures shape knowledge acquisition during pre-training, and how this process differs from dense architectures, remains unknown. To address this issue, we introduce Gated-LPI (Log-Probability Increase), a neuron-level attribution metric that decomposes log-probability increase across neurons. We present a time-resolved comparison of knowledge acquisition dynamics in MoE and dense architectures, tracking checkpoints over 1.2M training steps (~ 5.0T tokens) and 600K training steps (~ 2.5T tokens), respectively. Our experiments uncover three patterns: (1) Low-entropy backbone. The top approximately 1% of MoE neurons capture over 45% of positive updates, forming a high-utility core, which is absent in the dense baseline. (2) Early consolidation. The MoE model locks into a stable importance profile within < 100K steps, whereas the dense model remains volatile throughout training. (3) Functional robustness. Masking the ten most important MoE attention heads reduces relational HIT@10 by < 10%, compared with > 50% for the dense model, showing that sparsity fosters distributed -- rather than brittle -- knowledge storage. These patterns collectively demonstrate that sparsity fosters an intrinsically stable and distributed computational backbone from early in training, helping bridge the gap between sparse architectures and training-time interpretability.

cs.AI↗

arXiv2026

Optical Quasi-symmetry Groups for Meron Lattices

We introduce quasi-symmetry groups in optics emerging from the commutation between mirror operation and the spin-orbit interaction (SOI) of light. Contrary to the principle of symmetry inheritance in free-space optics, where the symmetry of any structured field is strictly constrained by that of its source, we show that strong SOI enables quasi-symmetry-protected formation of meron lattices even when the underlying optical sources violate the nominal rotational symmetry. By analyzing the Hermiticity of the electric-dipole radiation amplitude in a circular polarization basis, we derive an effective mirror operator acting only on a subset of C3 polarized dipole emitters, forming a quasi-symmetry group that commutes with SOI. This quasi-symmetry guarantees exact C3 merons and gives rise to a robust polarization zone within which continuously varying input polarizations generate identical topological textures. Our work establishes quasi-symmetry as a new fundamental principle in optical physics and opens pathways to engineered topological structures of light beyond conventional symmetry constraints.

physics.optics↗

arXiv2026

Multi-hop Reasoning via Early Knowledge Alignment

Retrieval-Augmented Generation (RAG) has emerged as a powerful paradigm for Large Language Models (LLMs) to address knowledge-intensive queries requiring domain-specific or up-to-date information. To handle complex multi-hop questions that are challenging for single-step retrieval, iterative RAG approaches incorporating reinforcement learning have been proposed. However, existing iterative RAG systems typically plan to decompose questions without leveraging information about the available retrieval corpus, leading to inefficient retrieval and reasoning chains that cascade into suboptimal performance. In this paper, we introduce Early Knowledge Alignment (EKA), a simple but effective module that aligns LLMs with retrieval set before planning in iterative RAG systems with contextually relevant retrieved knowledge. Extensive experiments on six standard RAG datasets demonstrate that by establishing a stronger reasoning foundation, EKA significantly improves retrieval precision, reduces cascading errors, and enhances both performance and efficiency. Our analysis from an entropy perspective demonstrate that incorporating early knowledge reduces unnecessary exploration during the reasoning process, enabling the model to focus more effectively on relevant information subsets. Moreover, EKA proves effective as a versatile, training-free inference strategy that scales seamlessly to large models. Generalization tests across diverse datasets and retrieval corpora confirm the robustness of our approach. Overall, EKA advances the state-of-the-art in iterative RAG systems while illuminating the critical interplay between structured reasoning and efficient exploration in reinforcement learning-augmented frameworks. The code is released at \href{https://github.com/yxzwang/EarlyKnowledgeAlignment}{Github}.

cs.CL↗

arXiv2025

The Formation of Electron-capture Supernovae: A Review

It is generally believed that the electron-capture reactions happen when the oxygen-neon (ONe) cores grow in masses close to the Chandrasekhar limit, leading to the formation of neutron stars (NSs) via electron-capture supernovae (EC-SNe). EC-SNe are predicted to be the most likely short-lived and faint optical transients, and a small ejecta mass is expected during the collapse. This kind of SNe provide an alternative channel for producing isolated NSs and NS systems, especially for the formation of X-ray binaries and double NSs. Although EC-SNe were proposed ~45 yr ago, there are still some uncertainties for the origin of EC-SNe. In this article, we review recent studies on the two classic progenitor channels of EC-SNe, i.e., the single star channel and the binary star channel. In the single star channel, EC-SNe can happen in super asymptotic giant branch stars or He stars, whereas in the binary star channel EC-SNe can occur in He stars in binaries (involving He star+MS systems and NS+He star systems) or accretion-induced collapse in white dwarf binaries (involving the single-degenerate scenario and the double-degenerate scenario). Recent progress on the two progenitor channels is discussed, including the initial parameter range for EC-SNe, the evolutionary paths to EC-SNe, related objects and some observational constraints, etc. We also make some discussions on the possible candidates for EC-SNe in this article, and the impacts of EC-SNe on some research fields, e.g., the properties of NSs, double NS population and chemical products, etc. We also discuss the differences between EC-SNe and ultra-stripped SNe in this article. Research on EC-SNe is at a pivotal stage, with key theoretical uncertainties and observational challenges requiring integrated modeling and multi-wavelength observations for robust identification.

astro-ph.HE↗

arXiv2025

Intersecting well approximable and missing digit sets

Let $b\geq3$ be an integer and $C(b,D)$ be the set of real numbers in $[0,1]$ whose $b$-ary expansion consists of digits restricted to a given set $D\subseteq\{0,\ldots,b-1\}$. Given an integer $t\geq2$ and a real, positive function $ψ$, let $W_{t}(ψ)$ denote the set of $x$ in $[0,1]$ for which $|x-p/t^{n}|<ψ(n)$ for infinitely many $(p,n)\in\mathbb{Z}\times\mathbb{N}$. We prove a general Hausdorff dimension result concerning the intersection of $W_{t}(ψ)$ with an arbitrary self similar set which implies that $\dim_{\rm H}(W_{t}(ψ)\cap C(b,D))\le\dim_{\rm H}W_{t}(ψ)\times \dim_{\rm H}C(b,D)$. When $b$ and $t$ have the same prime divisors, under certain restrictions on the digit set $D$, we give a sufficient condition for the Hausdorff measure of $W_{t}(ψ)\cap C(b,D)$ to be zero. This closes a gap in a result of Li, Li and Wu \cite{LLW2025} and shows that the dimension of the intersection can be strictly less than the product of the dimensions. The latter disproves the product conjecture of Li, Li and Wu.

math.NT↗

arXiv2025

Grad: Guided Relation Diffusion Generation for Graph Augmentation in Graph Fraud Detection

Nowadays, Graph Fraud Detection (GFD) in financial scenarios has become an urgent research topic to protect online payment security. However, as organized crime groups are becoming more professional in real-world scenarios, fraudsters are employing more sophisticated camouflage strategies. Specifically, fraudsters disguise themselves by mimicking the behavioral data collected by platforms, ensuring that their key characteristics are consistent with those of benign users to a high degree, which we call Adaptive Camouflage. Consequently, this narrows the differences in behavioral traits between them and benign users within the platform's database, thereby making current GFD models lose efficiency. To address this problem, we propose a relation diffusion-based graph augmentation model Grad. In detail, Grad leverages a supervised graph contrastive learning module to enhance the fraud-benign difference and employs a guided relation diffusion generator to generate auxiliary homophilic relations from scratch. Based on these, weak fraudulent signals would be enhanced during the aggregation process, thus being obvious enough to be captured. Extensive experiments have been conducted on two real-world datasets provided by WeChat Pay, one of the largest online payment platforms with billions of users, and three public datasets. The results show that our proposed model Grad outperforms SOTA methods in both various scenarios, achieving at most 11.10% and 43.95% increases in AUC and AP, respectively. Our code is released at https://github.com/AI4Risk/antifraud and https://github.com/Muyiiiii/WWW25-Grad.

cs.LG↗

arXiv2025

Evolution Models of CO WD -- AGB Star Merger Remnants

Common envelope evolution is a critical but still poorly understood phase in binary evolution. It plays a key role in forming close binaries such as hot subdwarfs, double white dwarfs, X-ray binaries, and double neutron stars. However, its outcomes remain highly uncertain. Depending on the efficiency of envelope ejection, a system may either survive as a close binary or undergo a complete merger. In this work, we investigate the post merger evolution of systems where a CO WD mergers with the core of an AGB star. A grid of merger remnant models with various core and envelope masses is constructed. At the onset of evolution, the CO core contracts and undergoes off-center carbon ignition, producing an inwardly propagating carbon flame. For remnants with relatively low mass of CO core, the flame phase is followed by core contraction and subsequent H-shell burning. For more massive CO cores, the carbon flame reaches the center and is soon followed by off-center neon burning, which is expected to eventually lead to core-collapse supernovae. The merger remnants occupy nearly the same region on HR diagram as ordinary AGB or super-AGB stars, exhibiting similar surface properties. Although their surface abundance may differ slightly from those of normal AGB stars depending on the initial core and envelope masses, these differences are strongly reduced once mass-loss is taken into account. We suggest that some giant-like stars, including candidates for Thorne-Zytkow objects (e.g., HV 2112), might alternatively be explained as AGB-WD merger remnants.

astro-ph.SR↗

arXiv2025

TF-MCL: Time-frequency Fusion and Multi-domain Cross-Loss for Self-supervised Depression Detection

In recent years, there has been a notable increase in the use of supervised detection methods of major depressive disorder (MDD) based on electroencephalogram (EEG) signals. However, the process of labeling MDD remains challenging. As a self-supervised learning method, contrastive learning could address the shortcomings of supervised learning methods, which are unduly reliant on labels in the context of MDD detection. However, existing contrastive learning methods are not specifically designed to characterize the time-frequency distribution of EEG signals, and their capacity to acquire low-semantic data representations is still inadequate for MDD detection tasks. To address the problem of contrastive learning method, we propose a time-frequency fusion and multi-domain cross-loss (TF-MCL) model for MDD detection. TF-MCL generates time-frequency hybrid representations through the use of a fusion mapping head (FMH), which efficiently remaps time-frequency domain information to the fusion domain, and thus can effectively enhance the model's capacity to synthesize time-frequency information. Moreover, by optimizing the multi-domain cross-loss function, the distribution of the representations in the time-frequency domain and the fusion domain is reconstructed, thereby improving the model's capacity to acquire fusion representations. We evaluated the performance of our model on the publicly available datasets MODMA and PRED+CT and show a significant improvement in accuracy, outperforming the existing state-of-the-art (SOTA) method by 5.87% and 9.96%, respectively.

cs.LG↗