arXiv subjects

Bo Lv

Publications and source records attributed to Bo Lv.

At least 19 recordsLinked to original sources

arXiv2026

Orthogonal Representation Editing: Decoupling Semantic Entanglement in Batch Knowledge Editing of LLMs

Knowledge editing aims to efficiently update factual information in Large Language Models (LLMs) without full retraining. However, existing methods still suffer from performance degradation in batch knowledge editing. We identify that semantic representation entanglement, such as overlapping concepts and shared syntactic patterns, accumulates interference in the representation space and reduces editing precision. To bridge this gap, in this paper, we propose Orthogonal Representation Editing (ORE), which performs edits in the hidden representation space of LLMs by constructing a general semantic subspace and enforcing orthogonal constraints on edit vectors, effectively decoupling semantic entanglement. Furthermore, we introduce a gated non-linear representation head to enable adaptive learning of editing locations and precise control over knowledge injection. Extensive experiments show that ORE outperforms existing methods and achieves superior performance in cross-lingual knowledge editing scenarios. We release our code at https://github.com/YVVH/ORE.

cs.CL↗

arXiv2026

Beyond Query Memorization: Large Language Model Routing with Query Decomposition and Historical Matching

Optimizing the trade-off among predictive performance and computational cost is a central focus in the deployment of Large Language Models (LLMs). Current routing methods primarily rely on direct mapping from queries to models based on surface-level features, making them susceptible to the memorization trap and leading to poor generalizability on out-of-distribution (OOD) data. In this paper, we propose DecoR, a novel routing framework that recasts the routing task as a matching process of sifting similar queries from historical logs, effectively mitigating the memorization trap. To enhance matching accuracy, we introduce a query capability deconstruction method that decouples linguistic surface forms from task-intrinsic requirements, directing matching toward capability dimensions to ground decisions in essential task attributes. Furthermore, we develop CodaSet, a comprehensive benchmark for assessing routing generalization, where experimental results demonstrate that DecoR maintains superior accuracy while substantially lowering inference costs across both in-distribution and OOD settings. All the codes and data are available at https://github.com/lvbotenbest/DecoR.

cs.AI↗

arXiv2026

RouteScan: A Non-Intrusive Approach to Auditing MoE LLMs Safety via Expert Routing Telemetry

As Mixture-of-Experts (MoE) architectures are increasingly adopted for scaling Large Language Models (LLMs), safety auditing becomes necessary to verify whether these models produce or facilitate harmful behaviors during operation. However, existing content-based auditing methods typically require access to user prompts, model internals, or outputs, potentially exposing sensitive user information and creating a tension between LLM safety and user privacy. On the other hand, we observe that, in MoE models, different inputs induce different sparse expert-routing patterns, which produce measurable footprints in low-level GPU execution telemetry. We refer to these hardware-observable signals induced by expert-routing decisions as expert routing telemetry; they are derived from GPU execution rather than from router logits or token-level routing assignments. Inspired by this observation, we propose RouteScan, a non-intrusive auditing framework for detecting harmful behaviors through such routing-induced GPU telemetry. Specifically, RouteScan utilizes the number of active GPU threads allocated to expert modules during the prefilling phase as a discriminative micro-architectural fingerprint, and builds a lightweight detection pipeline that isolates cross-domain invariant risk indicators for the precise identification of malicious prompts. Comprehensive evaluations on four open-source MoE LLMs with distinct routing designs demonstrate that RouteScan achieves strong generalization, with an AUROC exceeding 0.91 on unseen harmful domains. Moreover, privacy stress tests show that, although aggregated execution telemetry retains input-related attribute information, full prompts and exact sensitive fields cannot be reliably recovered under the evaluated attacks.

cs.CR↗

arXiv2026

Hy-MT2: A Family of Fast, Efficient and Powerful Multilingual Translation Models in the Wild

Hy-MT2 is a family of fast-thinking multilingual translation models designed for complex real-world scenarios. It includes three model sizes: 1.8B, 7B, and 30B-A3B (MoE), all of which support translation among 33 languages and effectively follow translation instructions in multiple languages. Multi-dimensional evaluations show that Hy-MT2 delivers outstanding performance across general, real-world business, domain-specific, and instruction-following translation tasks. The 7B and 30B models outperform open-source models such as DeepSeek-V4-Pro and Kimi K2.6 in fast-thinking mode, while the lightweight 1.8B model also surpasses mainstream commercial APIs from providers such as Microsoft and Doubao overall. Moreover, when paired with AngelSlim's 1.25-bit extreme quantization for on-device deployment, the lightweight 1.8B model requires only 440 MB of storage and achieves a 1.5x inference speedup.

cs.CL↗

arXiv2026

MOGeo: Beyond One-to-One Cross-View Object Geo-localization

Cross-View Object Geo-Localization (CVOGL) aims to locate an object of interest in a query image within a corresponding satellite image. Existing methods typically assume that the query image contains only a single object, which does not align with the complex, multi-object geo-localization requirements in real-world applications, making them unsuitable for practical scenarios. To bridge the gap between the realistic setting and existing task, we propose a new task, called Cross-View Multi-Object Geo-Localization (CVMOGL). To advance the CVMOGL task, we first construct a benchmark, CMLocation, which includes two datasets: CMLocation-V1 and CMLocation-V2. Furthermore, we propose a novel cross-view multi-object geo-localization method, MOGeo, and benchmark it against existing state-of-the-art methods. Extensive experiments are conducted under various application scenarios to validate the effectiveness of our method. The results demonstrate that cross-view object geo-localization in the more realistic setting remains a challenging problem, encouraging further research in this area.

cs.CV↗

arXiv2026

OneLatent: Single-Token Compression for Visual Latent Reasoning

Chain-of-thought (CoT) prompting improves reasoning but often increases inference cost by one to two orders of magnitude. To address these challenges, we present \textbf{OneLatent}, a framework that compresses intermediate reasoning into a single latent token via supervision from rendered CoT images and DeepSeek-OCR hidden states. By rendering textual steps into images, we obtain a deterministic supervision signal that can be inspected and audited without requiring the model to output verbose textual rationales. Across benchmarks, OneLatent reduces average output length by $11\times$ with only a $2.21\%$ average accuracy drop relative to textual CoT, while improving output token contribution (OTC) by $6.8\times$. On long-chain logical reasoning, OneLatent reaches $99.80\%$ on ProntoQA and $97.80\%$ on ProsQA with one latent token, with compression up to $87.4\times$, supporting compression-constrained generalization.

cs.AI↗

arXiv2025

From Topology to Behavioral Semantics: Enhancing BGP Security by Understanding BGP's Language with LLMs

The trust-based nature of Border Gateway Protocol (BGP) makes it vulnerable to disruptions like prefix hijacking and misconfigurations, threatening routing stability. Traditional detection relies on manual inspection with limited scalability. Machine/Deep Learning (M/DL) approaches automate detection but suffer from suboptimal precision, limited generalizability, and high retraining costs. This is because existing methods focus on topological structures rather than comprehensive semantic characteristics of Autonomous Systems (ASes), often misinterpreting functionally similar but topologically distant ASes. To address this, we propose BGPShield, an anomaly detection framework built on LLM embeddings that captures the Behavior Portrait and Routing Policy Rationale of each AS beyond topology, such as operational scale and global role. We propose a segment-wise aggregation scheme to transform AS descriptions into LLM representations without information loss, and a lightweight contrastive reduction network to compress them into a semantic-consistent version. Using these representations, our AR-DTW algorithm aligns and accumulates semantic distances to reveal behavioral inconsistencies. Evaluated on 16 real-world datasets, BGPShield detects 100% of verified anomalies with a false discovery rate below 5%. Notably, the employed LLMs were released prior to evaluation events, verifying generalizability. Furthermore, BGPShield constructs representations for unseen ASes within one second, significantly outperforming BEAM which demands costly retraining (averaging 65 hours).

cs.NI↗

arXiv2025

HyCoRA: Hyper-Contrastive Role-Adaptive Learning for Role-Playing

Multi-character role-playing aims to equip models with the capability to simulate diverse roles. Existing methods either use one shared parameterized module across all roles or assign a separate parameterized module to each role. However, the role-shared module may ignore distinct traits of each role, weakening personality learning, while the role-specific module may overlook shared traits across multiple roles, hindering commonality modeling. In this paper, we propose a novel HyCoRA: Hyper-Contrastive Role-Adaptive learning framework, which efficiently improves multi-character role-playing ability by balancing the learning of distinct and shared traits. Specifically, we propose a Hyper-Half Low-Rank Adaptation structure, where one half is a role-specific module generated by a lightweight hyper-network, and the other half is a trainable role-shared module. The role-specific module is devised to represent distinct persona signatures, while the role-shared module serves to capture common traits. Moreover, to better reflect distinct personalities across different roles, we design a hyper-contrastive learning mechanism to help the hyper-network distinguish their unique characteristics. Extensive experimental results on both English and Chinese available benchmarks demonstrate the superiority of our framework. Further GPT-4 evaluations and visual analyses also verify the capability of HyCoRA to capture role characteristics.

cs.CL↗

arXiv2025

A Comprehensive Analysis of Churn Prediction in Telecommunications Using Machine Learning

Customer churn prediction in the telecommunications sector represents a critical business intelligence task that has evolved from subjective human assessment to sophisticated algorithmic approaches. In this work, we present a comprehensive framework for telecommunications churn prediction leveraging deep neural networks. Through systematic problem formulation, rigorous dataset analysis, and careful feature engineering, we develop a model that captures complex patterns in customer behavior indicative of potential churn. We conduct extensive empirical evaluations across multiple performance metrics, demonstrating that our proposed neural architecture achieves significant improvements over existing baseline methods. Our approach not only advances the state-of-the-art in churn prediction accuracy but also provides interpretable insights into the key factors driving customer attrition in telecommunications services.

stat.AP↗

arXiv2025

Realization of a non-Hermitian Haldane model in circuits

The Haldane model is the simplest yet most powerful topological lattice model exhibiting various phases, including the Dirac semimetal phase and the anomalous quantum Hall phase (also known as the Chern insulator). Although considered unlikely to be physically directly realizable in condensed matter systems, it has been experimentally demonstrated in other physical settings such as cold atoms, where Hermiticity is usually preserved. Extending this model to the non-Hermitian regime with energy non-conservation can significantly enrich topological phases that lack Hermitian counterparts; however, such exploration remains experimentally challenging due to the lack of suitable physical platforms. Here, based on electric circuits, we report the experimental realization of a genuine non-Hermitian Haldane model with asymmetric next-nearest-neighbor hopping. We observe two previously uncovered phases: a non-Hermitian Chern insulator and a non-Hermitian semimetal phase, both exhibiting boundary-dependent amplifying or dissipative chiral edge states. Our work paves the way for exploring non-Hermiticity-induced unconventional topological phases in the Haldane model.

physics.optics↗

arXiv2025

An Improved Pure Fully Connected Neural Network for Rice Grain Classification

Rice is a staple food for a significant portion of the world's population, providing essential nutrients and serving as a versatile in-gredient in a wide range of culinary traditions. Recently, the use of deep learning has enabled automated classification of rice, im-proving accuracy and efficiency. However, classical models based on first-stage training may face difficulties in distinguishing between rice varieties with similar external characteristics, thus leading to misclassifications. Considering the transparency and feasibility of model, we selected and gradually improved pure fully connected neural network to achieve classification of rice grain. The dataset we used contains both global and domestic rice images obtained from websites and laboratories respectively. First, the training mode was changed from one-stage training to two-stage training, which significantly contributes to distinguishing two similar types of rice. Secondly, the preprocessing method was changed from random tilting to horizontal or vertical position cor-rection. After those two enhancements, the accuracy of our model increased notably from 97% to 99%. In summary, two subtle methods proposed in this study can remarkably enhance the classification ability of deep learning models in terms of the classification of rice grain.

cs.CV↗

arXiv2025

GRAPHMOE: Amplifying Cognitive Depth of Mixture-of-Experts Network via Introducing Self-Rethinking Mechanism

Traditional Mixture-of-Experts (MoE) networks benefit from utilizing multiple smaller expert models as opposed to a single large network. However, these experts typically operate independently, leaving a question open about whether interconnecting these models could enhance the performance of MoE networks. In response, we introduce GRAPHMOE, a novel method aimed at augmenting the cognitive depth of language models via a self-rethinking mechanism constructed on Pseudo GraphMoE networks. GRAPHMOE employs a recurrent routing strategy to simulate iterative thinking steps, thereby facilitating the flow of information among expert nodes. We implement the GRAPHMOE architecture using Low-Rank Adaptation techniques (LoRA) and conduct extensive experiments on various benchmark datasets. The experimental results reveal that GRAPHMOE outperforms other LoRA based models, achieving state-of-the-art (SOTA) performance. Additionally, this study explores a novel recurrent routing strategy that may inspire further advancements in enhancing the reasoning capabilities of language models.

cs.CL↗

arXiv2024

SpecFuse: Ensembling Large Language Models via Next-Segment Prediction

Ensembles of generative large language models (LLMs) are a promising way to compensate for individual model limitations, integrating the strengths of different LLMs. Existing LLM ensemble methods, however, face limitations such as first-token delay and challenges in long-range semantic collaboration between models, Moreover, they typically assume equal voting weights for all models during ensemble, ignoring task-specific performance differences among models. In this work, we propose SpecEM, a training-free, plug-and-play LLM ensemble framework that dynamically adjusts each model's model contribution in real time based on task performance. Inspired by speculative decoding, SpecEM iteratively performs drafting and verification, allowing models to collaborate semantically at the segment level for integrated output. Furthermore, we introduce an online feedback mechanism with multiplicative weight updates, where each model's voting weight is adjusted on-the-fly according to how often it outperforms others during verification stage, ensuring that stronger models exert greater influence during ensembling. Experimental results on five LLM families (ranging from 7B to 72B parameters) and six benchmark datasets, spanning open-domain instruction following, reasoning, commonsense, demonstrate consistent performance improvements compared to state-of-the-art LLM ensemble methods. Our code is available at https://github.com/lvbotenbest/SpecEM.

cs.CL↗

arXiv2024

KVPruner: Structural Pruning for Faster and Memory-Efficient Large Language Models

The bottleneck associated with the key-value(KV) cache presents a significant challenge during the inference processes of large language models. While depth pruning accelerates inference, it requires extensive recovery training, which can take up to two weeks. On the other hand, width pruning retains much of the performance but offers slight speed gains. To tackle these challenges, we propose KVPruner to improve model efficiency while maintaining performance. Our method uses global perplexity-based analysis to determine the importance ratio for each block and provides multiple strategies to prune non-essential KV channels within blocks. Compared to the original model, KVPruner reduces runtime memory usage by 50% and boosts throughput by over 35%. Additionally, our method requires only two hours of LoRA fine-tuning on small datasets to recover most of the performance.

cs.CL↗

arXiv2024

BioMNER: A Dataset for Biomedical Method Entity Recognition

Named entity recognition (NER) stands as a fundamental and pivotal task within the realm of Natural Language Processing. Particularly within the domain of Biomedical Method NER, this task presents notable challenges, stemming from the continual influx of domain-specific terminologies in scholarly literature. Current research in Biomedical Method (BioMethod) NER suffers from a scarcity of resources, primarily attributed to the intricate nature of methodological concepts, which necessitate a profound understanding for precise delineation. In this study, we propose a novel dataset for biomedical method entity recognition, employing an automated BioMethod entity recognition and information retrieval system to assist human annotation. Furthermore, we comprehensively explore a range of conventional and contemporary open-domain NER methodologies, including the utilization of cutting-edge large-scale language models (LLMs) customised to our dataset. Our empirical findings reveal that the large parameter counts of language models surprisingly inhibit the effective assimilation of entity extraction patterns pertaining to biomedical methods. Remarkably, the approach, leveraging the modestly sized ALBERT model (only 11MB), in conjunction with conditional random fields (CRF), achieves state-of-the-art (SOTA) performance.

cs.CL↗

arXiv2024

Subharmonic oscillations in the Floquet circuit with the frequency-synthesis dimension

The period-doubling oscillation emerges with the coexistence between zero and {\pi} modes in Floquet topological insulator. Here, utilized the flexibility of the circuit, we construct the Floquet circuit with frequency-synthetic dimension and find the topological-protected deeply-subharmonic oscillations with the period extensively exceeding the doubling-driven period. In the construction framework, the periodically-driven mechanism is attained by implementing the circuit-oscillator hierarchy with the stepping-variation resonances in frequency domain. The zero and {\pi} modes that arise at the Floquet band in the circuit indicate the anomalous boundary-bulk correspondence. The coexistence of zero and {\pi} modes, results in a subharmonic oscillation with the extremely-low frequency on the edge of the Floquet circuit. Furthermore, we explore the Floquet band with the enhanced periodically-driven strength tailored by the component flexibility of the circuit. Our method provides a flexible scheme to study Floquet topological phases, and open a new path for realizing the deeply subwavelength system.

cond-mat.mes-hall↗

arXiv2022

Mistakes of A Popular Protocol Calculating Private Set Intersection and Union Cardinality and Its Corrections

In 2012, De Cristofaro et al. proposed a protocol to calculate the Private Set Intersection and Union cardinality(PSI-CA and PSU-CA). This protocol's security is based on the famous DDH assumption. Since its publication, it has gained lots of popularity because of its efficiency(linear complexity in computation and communication) and concision. So far, it's still considered one of the most efficient PSI-CA protocols and the most cited(more than 170 citations) PSI-CA paper based on the Google Scholar search. However, when we tried to implement this protocol, we couldn't get the correct result of the test data. Since the original paper lacks of experimental results to verify the protocol's correctness, we looked deeper into the protocol and found out it made a fundamental mistake. Needless to say, its correctness analysis and security proof are also wrong. In this paper, we will point out this PSI-CA protocol's mistakes, and provide the correct version of this protocol as well as the PSI protocol developed from this protocol. We also present a new security proof and some experimental results of the corrected protocol.

cs.CR↗

arXiv2021

Realization of quasicrystalline quadrupole topological insulators in electrical circuits

Quadrupole topological insulators are a new class of topological insulators with quantized quadrupole moments, which support protected gapless corner states. The experimental demonstrations of quadrupole-topological insulators were reported in a series of artificial materials, such as photonic crystals, acoustic crystals, and electrical circuits. In all these cases, the underlying structures have discrete translational symmetry and thus are periodic. Here we experimentally realize two-dimensional aperiodic-quasicrystalline quadrupole-topological insulators by constructing them in electrical circuits, and observe the spectrally and spatially localized corner modes. In measurement, the modes appear as topological boundary resonances in the corner impedance spectra. Additionally, we demonstrate the robustness of corner modes on the circuit. Our circuit design may be extended to study topological phases in higher-dimensional aperiodic structures.

cond-mat.mes-hall↗