arXiv subjects

Lang Zhou

Publications and source records attributed to Lang Zhou.

11 recordsLinked to original sources

arXiv2026

CMT-RAG: Complementary Memory Traces for Multi-turn Multi-hop RAG

Multi-turn information-seeking conversations require both multi-hop reasoning and long-range dependency tracking across turns. However, existing RAG systems typically represent conversational memory as raw dialogue history, rewritten queries, or unstructured summaries, making it difficult to recover the specific prior reasoning steps and evidence required for follow-up queries. Our key insight is to align conversational memory with retrieval by representing dialogue context as sub-question-level reasoning traces. Building on this insight, we introduce MuMu-QA, a benchmark for multi-turn multi-hop RAG with explicit cross-turn sub-question dependency annotations, and CMT-RAG, a complementary memory framework for this setting. At each turn, CMT-RAG employs a state-space trace generator, whose recurrent state serves as runtime memory, to incorporate recent conversational context and decompose the current query into structured trace drafts containing retrieval-oriented sub-questions and dependencies on earlier traces. It then grounds these drafts with retrieved evidence and stores them as persistent memory traces in a session-level DAG, enabling future turns to efficiently recover relevant prior reasoning and evidence. Experiments on MuMu-QA and corpus-level RAG benchmarks show that CMT-RAG consistently outperforms five categories of RAG baselines in answer accuracy.

cs.CL↗

arXiv2026

UT-ACA: Uncertainty-Triggered Adaptive Context Allocation for Long-Context Inference

Long-context inference remains challenging for large language models due to attention dilution and out-of-distribution degradation. Context selection mitigates this limitation by attending to a subset of key-value cache entries, yet most methods allocate a fixed context budget throughout decoding despite highly non-uniform token-level contextual demands. To address this issue, we propose Uncertainty-Triggered Adaptive Context Allocation (UT-ACA), an inference-time framework that dynamically adjusts the context window based on token-wise uncertainty. UT-ACA learns an uncertainty detector that combines semantic embeddings with logit-based confidence while accounting for uncertainty accumulation across decoding steps. When insufficient evidence is indicated, UT-ACA selectively rolls back, expands the context window, and regenerates the token with additional support. Experiments show that UT-ACA substantially reduces average context usage while preserving generation quality in long-context settings. Code and dataset are available at https://github.com/Tommy307/UT-ACA.

cs.CL↗

arXiv2026

ReViP: Mitigating False Completion in Vision-Language-Action Models with Vision-Proprioception Rebalance

Vision-Language-Action (VLA) models have advanced robotic manipulation by combining vision, language, and proprioception to predict actions. However, previous methods fuse proprioceptive signals directly with vision-language features, resulting in state-dominant bias and \textbf{false completions} despite visible execution failures. We systematically analyze this failure mode, attributing it to modality imbalance, where policies overly rely on internal state progression and underuse visual evidence. To address this, we introduce the first \textbf{False-Completion Benchmark Suite}, featuring eight tasks with three controlled perturbations (\emph{Object Drop}, \emph{Distractor Swap}, \emph{Relayout}) to comprehensively evaluate false completion. Moreover, we propose \textbf{ReViP}, a novel VLA framework with \textbf{Vi}sion-\textbf{P}roprioception \textbf{Re}balance to enhance visual grounding and robustness under perturbations. The key insight is to introduce auxiliary \emph{progress-aware visual cues} to adaptively modulate the coupling between semantic perception and proprioceptive dynamics. Specifically, progress-aware visual cues are extracted by an external Task-Stage Observer, which performs task-relevant reasoning on real-time observations to drive task-stage feature-wise linear modulation, enhancing environmental awareness and mitigating state-driven errors. Extensive experiments show that ReViP effectively mitigates false completion and improves success rates over strong VLA baselines, achieving a \textbf{26\%} gain over $π_0$ model on our suite, with gains extending to LIBERO, RoboTwin 2.0, and real-world evaluations.

cs.RO↗

arXiv2026

Mode Conversion of Hyperbolic Phonon Polaritons in van der Waals terraces

Electromagnetic hyperbolicity has driven key functionalities in nanophotonics, including super-resolution imaging, efficient energy control, and extreme light manipulation. Central to these advances are hyperbolic polaritons - nanometer-scale light-matter waves - spanning multiple energy-momentum dispersion orders with distinct mode profiles and incrementally high optical momenta. In this work, we report the mode conversion of hyperbolic polaritons across different dispersion orders by breaking the structure symmetry in engineered step-shape van der Waals (vdW) terraces. The mode conversion from the fundamental to high-order hyperbolic polaritons is imaged using scattering-type scanning near-field optical microscopy (s-SNOM) on both hexagonal boron nitride (hBN) and alpha-phase molybdenum trioxide (alpha-MoO3) vdW terraces. Our s-SNOM data, augmented with electromagnetic simulations, further demonstrate the alteration of polariton mode conversion by varying the step size of vdW terraces. The mode conversion reported here offers a practical approach toward integrating previously independent different-order hyperbolic polaritons with ultra-high momenta, paving the way for promising applications in nano-optical circuits, sensing, computation, information processing, and super-resolution imaging.

physics.optics↗

arXiv2025

Multilingual BERT language model for medical tasks: Evaluation on domain-specific adaptation and cross-linguality

In multilingual healthcare applications, the availability of domain-specific natural language processing(NLP) tools is limited, especially for low-resource languages. Although multilingual bidirectional encoder representations from transformers (BERT) offers a promising motivation to mitigate the language gap, the medical NLP tasks in low-resource languages are still underexplored. Therefore, this study investigates how further pre-training on domain-specific corpora affects model performance on medical tasks, focusing on three languages: Dutch, Romanian and Spanish. In terms of further pre-training, we conducted four experiments to create medical domain models. Then, these models were fine-tuned on three downstream tasks: Automated patient screening in Dutch clinical notes, named entity recognition in Romanian and Spanish clinical notes. Results show that domain adaptation significantly enhanced task performance. Furthermore, further differentiation of domains, e.g. clinical and general biomedical domains, resulted in diverse performances. The clinical domain-adapted model outperformed the more general biomedical domain-adapted model. Moreover, we observed evidence of cross-lingual transferability. Moreover, we also conducted further investigations to explore potential reasons contributing to these performance differences. These findings highlight the feasibility of domain adaptation and cross-lingual ability in medical NLP. Within the low-resource language settings, these findings can provide meaningful guidance for developing multilingual medical NLP systems to mitigate the lack of training data and thereby improve the model performance.

cs.CL↗

arXiv2025

Automated HIV Screening on Dutch Electronic Health Records with Large Language Models

Efficient screening and early diagnosis of HIV are critical for reducing onward transmission. Although large scale laboratory testing is not feasible, the widespread adoption of Electronic Health Records (EHRs) offers new opportunities to address this challenge. Existing research primarily focuses on applying machine learning methods to structured data, such as patient demographics, for improving HIV diagnosis. However, these approaches often overlook unstructured text data such as clinical notes, which potentially contain valuable information relevant to HIV risk. In this study, we propose a novel pipeline that leverages a Large Language Model (LLM) to analyze unstructured EHR text and determine a patient's eligibility for further HIV testing. Experimental results on clinical data from Erasmus University Medical Center Rotterdam demonstrate that our pipeline achieved high accuracy while maintaining a low false negative rate.

cs.CL↗

arXiv2025

Cross-Lingual Consistency: A Novel Inference Framework for Advancing Reasoning in Large Language Models

Chain-of-thought (CoT) has emerged as a critical mechanism for enhancing reasoning capabilities in large language models (LLMs), with self-consistency demonstrating notable promise in boosting performance. However, inherent linguistic biases in multilingual training corpora frequently cause semantic drift and logical inconsistencies, especially in sub-10B parameter LLMs handling complex inference tasks. To overcome these constraints, we propose the Cross-Lingual Consistency (CLC) framework, an innovative inference paradigm that integrates multilingual reasoning paths through majority voting to elevate LLMs' reasoning capabilities. Empirical evaluations on the CMATH dataset reveal CLC's superiority over the conventional self-consistency method, delivering 9.5%, 6.5%, and 6.0% absolute accuracy gains for DeepSeek-Math-7B-Instruct, Qwen2.5-Math-7B-Instruct, and Gemma2-9B-Instruct respectively. Expanding CLC's linguistic scope to 11 diverse languages implies two synergistic benefits: 1) neutralizing linguistic biases in multilingual training corpora through multilingual ensemble voting, 2) escaping monolingual reasoning traps by exploring the broader multilingual solution space. This dual benefits empirically enables more globally optimal reasoning paths compared to monolingual self-consistency baselines, as evidenced by the 4.1%-18.5% accuracy gains using Gemma2-9B-Instruct on the MGSM dataset.

cs.CL↗

arXiv2024

Underdetermined DOA Estimation of Off-Grid Sources Based on the Generalized Double Pareto Prior

In this letter, we investigate a new generalized double Pareto based on off-grid sparse Bayesian learning (GDPOGSBL) approach to improve the performance of direction of arrival (DOA) estimation in underdetermined scenarios. The method aims to enhance the sparsity of source signal by utilizing the generalized double Pareto (GDP) prior. Firstly, we employ a first-order linear Taylor expansion to model the real array manifold matrix, and Bayesian inference is utilized to calculate the off-grid error, which mitigates the grid dictionary mismatch problem in underdetermined scenarios. Secondly, an innovative grid refinement method is introduced, treating grid points as iterative parameters to minimize the modeling error between the source and grid points. The numerical simulation results verify the superiority of the proposed strategy, especially when dealing with a coarse grid and few snapshots.

eess.SP↗

arXiv2022

Mars Rover Localization Based on A2G Obstacle Distribution Pattern Matching

Rover localization is one of the perquisites for large scale rover exploration. In NASA's Mars 2020 mission, the Ingenuity helicopter is carried together with the rover, which is capable of obtaining high-resolution imagery of Mars terrain, and it is possible to perform localization based on aerial-to-ground (A2G) imagery correspondence. However, considering the low-texture nature of the Mars terrain, and large perspective changes between UAV and rover imagery, traditional image matching methods will struggle to obtain valid image correspondence. In this paper we propose a novel pipeline for Mars rover localization. An algorithm combing image-based rock detection and rock distribution pattern matching is used to acquire A2G imagery correspondence, thus establishing the rover position in a UAV-generated ground map. Feasibility of this method is evaluated on sample data from a Mars analogue environment. The proposed method can serve as a reliable assist in future Mars missions.

cs.CV↗

arXiv2020

Deep Double-Side Learning Ensemble Model for Few-Shot Parkinson Speech Recognition

Diagnosis and therapeutic effect assessment of Parkinson disease based on voice data are very important,but its few-shot learning problem is challenging.Although deep learning is good at automatic feature extraction, it suffers from few-shot learning problem. Therefore, the general effective method is first conduct feature extraction based on prior knowledge, and then carry out feature reduction for subsequent classification. However, there are two major problems: 1) Structural information among speech features has not been mined and new features of higher quality have not been reconstructed. 2) Structural information between data samples has not been mined and new samples with higher quality have not been reconstructed. To solve these two problems, based on the existing Parkinson speech feature data set, a deep double-side learning ensemble model is designed in this paper that can reconstruct speech features and samples deeply and simultaneously. As to feature reconstruction, an embedded deep stacked group sparse auto-encoder is designed in this paper to conduct nonlinear feature transformation, so as to acquire new high-level deep features, and then the deep features are fused with original speech features by L1 regularization feature selection method. As to speech sample reconstruction, a deep sample learning algorithm is designed in this paper based on iterative mean clustering to conduct samples transformation, so as to obtain new high-level deep samples. Finally, the bagging ensemble learning mode is adopted to fuse the deep feature learning algorithm and the deep samples learning algorithm together, thereby constructing a deep double-side learning ensemble model. At the end of this paper, two representative speech datasets of Parkinson's disease were used for verification. The experimental results show that the proposed algorithm are effective.

cs.CV↗

arXiv2019

From planar junction to local junction: A new structure design of amorphous/crystalline silicon heterojunction solar cells for high efficiency and low cost

In order to obtain higher conversion efficiency and to reduce production cost for amorphous silicon/crystalline silicon (a-Si/c-Si) based heterojunction solar cells, a Heterojunction of Amorphous silicon and Crystalline silicon with Localized pn structure (HACL) has been designed. The potential performance of the HACL solar cell has been assessed by ATLAS simulation program. Its potential performance is compared with that of the Heterojunction with Intrinsic Thin film (HIT) and Heterojunction of Amorphous silicon and Crystalline silicon with Diffused junction (HACD) solar cells. The simulated results indicated that the conversion efficiency and the short-circuit current density of the HACL cell can reach to 28.18% and 43.06 mA/cm2, respectively, and are higher than that of the HIT and HACD cells. The main reasons for the great improvement are (1) to increase the light utilization rate on both sides of the cell and (2) to enhance the collection efficiency of the photocarriers. Moreover, the HACL structure can decrease the consumption of rare materials, such as indium, since the transparent conductive oxide (TCO) can be free in this structure. It is concluded that the HACL solar cell is a promising structure for high efficiency and low cost.

physics.app-ph↗