BZPEERPapers, people, evidence and research paths
Menu

Discover

SearchFind papers, people, topics, institutions and opportunities from one place.GraphSee how papers, authors, topics and evidence connect.

Research tools

HomeContinue from the papers, people and topics that matter now.CollectionsTurn saved papers into reading lists, evidence maps and shareable context.ResearchReconstruct the path that led you from search to paper to question.CollaborationMove from a paper, topic or expert request into a focused thread.PublishCreate full-text research articles with metadata, formulas and versions.

Network

ExpertsFind specialists whose work and reviews explain why they can help.CommunitiesJoin research groups organized around papers, topics and shared evidence.

Opportunities

ListingsBrowse roles, calls and collaborations connected to real research context.

Account

Log inReturn to your saved papers, graph views and active threads.Create accountStart saving papers, building a research trail and joining teams.
Log inCreate account

Source author record

Hua Wei

Hua Wei appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.

ResearcherUnclaimed source record
Artificial IntelligenceMachine LearningComputation and Languagecond-mat.mtrl-scimath.OCphysics.opticscond-mat.mes-hallcond-mat.str-elquant-ph

Catalog footprint

What is connected

13works
9topics
4close collaborators

Actions

Connect this record

Log in to claim
Open graphBrowse works

Research graph

See the researcher in context

Open full explorer

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

Published work

13 published item(s)

preprint2026arXiv

Agentic AI for Trip Planning Optimization Application

Trip planning for intelligent vehicles increasingly requires selecting optimal routes rather than merely producing feasible itineraries, as interacting factors such as travel time, energy consumption, and traffic conditions directly affect plan quality. Yet existing systems are largely designed for feasibility-oriented planning, and current benchmarks provide only reference answers without ground truth, preventing objective evaluation of optimization performance. In our paper, we address these limitations with an agentic AI framework that enables dynamic refinement through an orchestration agent coordinating specialized agents for traffic, charging, and points of interest, and with the Trip-planning Optimization Problems Dataset, which supplies definitive optimal solutions and category-level task structure for fine-grained analysis. Experiments show that our system achieves 77.4\% accuracy on the TOP Benchmark, significantly outperforming single-agent and workflow-based multi-agent baselines, demonstrating the importance of orchestrated agentic reasoning for robust trip planning optimization.

0 citations0 reviewsNo code linkedOpen access
preprint2026arXiv

Position: Uncertainty Quantification in LLMs is Just Unsupervised Clustering

Uncertainty Quantification (UQ) is widely regarded as the primary safeguard for deploying Large Language Models (LLMs) in high-stakes domains. However, we argue that the field suffers from a category error: mainstream UQ methods for LLMs are just unsupervised clustering algorithms. We demonstrate that most current approaches inherently quantify the internal consistency of the model's generations rather than their external correctness. Consequently, current methods are fundamentally blind to factual reality and fail to detect ``confident hallucinations,'' where models exhibit high confidence in stable but incorrect answers. Therefore, the current UQ methods may create a deceptive sense of safety when deploying the models with uncertainty. In detail, we identify three critical pathologies resulting from this dependence on internal state: a hyperparameter sensitivity crisis that renders deployment unsafe, an internal evaluation cycle that conflates stability with truth, and a fundamental lack of ground truth that forces reliance on unstable proxy metrics to evaluate uncertainty. To resolve this impasse, we advocate for a paradigm shift to UQ and outline a roadmap for the research community to adopt better evaluation metrics and settings, implement mechanism changes for native uncertainty, and anchor verification in objective truth, ensuring that model confidence serves as a reliable proxy for reality.

0 citations0 reviewsNo code linkedOpen access
preprint2024arXiv

Uncertainty Regularized Evidential Regression

The Evidential Regression Network (ERN) represents a novel approach that integrates deep learning with Dempster-Shafer's theory to predict a target and quantify the associated uncertainty. Guided by the underlying theory, specific activation functions must be employed to enforce non-negative values, which is a constraint that compromises model performance by limiting its ability to learn from all samples. This paper provides a theoretical analysis of this limitation and introduces an improvement to overcome it. Initially, we define the region where the models can't effectively learn from the samples. Following this, we thoroughly analyze the ERN and investigate this constraint. Leveraging the insights from our analysis, we address the limitation by introducing a novel regularization term that empowers the ERN to learn from the whole training set. Our extensive experiments substantiate our theoretical findings and demonstrate the effectiveness of the proposed solution.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

Semi-Supervised Clustering with Contrastive Learning for Discovering New Intents

Most dialogue systems in real world rely on predefined intents and answers for QA service, so discovering potential intents from large corpus previously is really important for building such dialogue services. Considering that most scenarios have few intents known already and most intents waiting to be discovered, we focus on semi-supervised text clustering and try to make the proposed method benefit from labeled samples for better overall clustering performance. In this paper, we propose Deep Contrastive Semi-supervised Clustering (DCSC), which aims to cluster text samples in a semi-supervised way and provide grouped intents to operation staff. To make DCSC fully utilize the limited known intents, we propose a two-stage training procedure for DCSC, in which DCSC will be trained on both labeled samples and unlabeled samples, and achieve better text representation and clustering performance. We conduct experiments on two public datasets to compare our model with several popular methods, and the results show DCSC achieve best performance across all datasets and circumstances, indicating the effect of the improvements in our work.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

A Probabilistic Simulator of Spatial Demand for Product Allocation

Connecting consumers with relevant products is a very important problem in both online and offline commerce. In physical retail, product placement is an effective way to connect consumers with products. However, selecting product locations within a store can be a tedious process. Moreover, learning important spatial patterns in offline retail is challenging due to the scarcity of data and the high cost of exploration and experimentation in the physical world. To address these challenges, we propose a stochastic model of spatial demand in physical retail. We show that the proposed model is more predictive of demand than existing baselines. We also perform a preliminary study into different automation techniques and show that an optimal product allocation policy can be learned through Deep Q-Learning.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

A Survey on Traffic Signal Control Methods

Traffic signal control is an important and challenging real-world problem, which aims to minimize the travel time of vehicles by coordinating their movements at the road intersections. Current traffic signal control systems in use still rely heavily on oversimplified information and rule-based methods, although we now have richer data, more computing power and advanced methods to drive the development of intelligent transportation. With the growing interest in intelligent transportation using machine learning methods like reinforcement learning, this survey covers the widely acknowledged transportation approaches and a comprehensive list of recent literature on reinforcement for traffic signal control. We hope this survey can foster interdisciplinary research on this important topic.

0 citations0 reviewsNo code linkedOpen access
preprint2016arXiv

An SQP Method Combined with Gradient Sampling for Small-Signal Stability Constrained OPF

Small-Signal Stability Constrained Optimal Power Flow (SSSC-OPF) can provide additional stability measures and control strategies to guarantee the system to be small-signal stable. However, due to the nonsmooth property of the spectral abscissa function, existing algorithms solving SSSC-OPF cannot guarantee convergence. To tackle this computational challenge of SSSC-OPF, we propose a Sequential Quadratic Programming (SQP) method combined with Gradient Sampling (GS) for SSSCOPF.At each iteration of the proposed SQP, the gradient of the spectral abscissa unction is randomly sampled at the current iterate and additional nearby points to make the search direction computation effective in nonsmooth regions. The method can guarantee SSSC-OPF is globally and efficiently convergent to stationary points with probability one. The effectiveness of the proposed method is tested and validated on WSCC 3-machine 9-bus system, New England 10-machine 39-bus system, and IEEE 54-machine 118-bus system.

0 citations0 reviewsNo code linkedOpen access
preprint2016arXiv

EV Dispatch Control for Supplementary Frequency Regulation Considering the Expectation of EV Owners

Electric Vehicles (EVs) are promising to provide frequency regulation services due to their fast regulating characteristics. However, when EVs participate in Supplementary Frequency Regulation (SFR), it is challenging to simultaneously achieve the dispatch of the control center and the expected State of Charge (SOC) levels of EV batteries. To solve this problem, in this paper we propose a Vehicle-to-Grid (V2G) control strategy, in which an uncertain dispatch is implemented in the control center without detailed EV charging/discharging information. The regulation from the control center is achieved by allocating the regulation task within the frequency regulation capacity (FRC) of EVs. The expected SOC levels of EV batteries are guaranteed by a real-time correction of their scheduled V2G power in EV charging stations. Simulations on an interconnected two-area power system validate the effectiveness of the proposed V2G control in achieving both the frequency regulation and the expected SOC levels of EVs.

0 citations0 reviewsNo code linkedOpen access
preprint2016arXiv

Magnetization reversal and negative volume thermal expansion in Fe doped Ca2RuO4

We report physical and structural properties of single-crystal Ca2Ru1-xFexO4 (0<x<0.20) as functions of temperature, magnetic field and pressure. Ca2RuO4 is a structurally-driven Mott insulator with a metal-insulator (MI) transition at TMI = 357 K, which is well separated from antiferromagnetic order at TN = 110 K. Fe substitution for Ru in Ca2RuO4 causes a pronounced magnetization reversal and giant negative volume thermal expansion (NVTE). The magnetization reversal is a result of a field-induced antiferromagnetic coupling between the Ru- and Fe-magnetic sublattices that have different temperature dependence. The NVTE is closely associated with the orthorhombic distortion, and becomes smaller as the orthorhombicity weakens due to either Fe doping or application of pressure. The study highlights an intriguing interplay between lattice, orbital and spin degrees of freedom that is at the root of the novel phenomena in Ca2RuO4.

0 citations0 reviewsNo code linkedOpen access
preprint2016arXiv

Pressure tuning the lattice and optical response of silver sulfide

Binary transition metal chalcogenides have attracted increasing attention for their unique structural and electronic properties. High pressure is powerful tool for tuning their crystal structure and electronic structure away from their pristine states.In this work, we systematically studied the in situ structural and optical behavior of silver sulfide (Ag2S) under pressure by X-ray diffraction (XRD) and Infrared (IR) spectroscopy measurements in a diamond anvil cell. Upon compression, Ag2S undergoes structural symmetrization from monoclinic to orthorhombic, represented by the decrease of angle from 99 to 90 degrees through a series of structural transitions coupled with lattice contractions. IR transmission and reflectivity measurements showed that pressure effectively tunes semiconducting Ag2S into a metal at ~ 22 GPa. Drude model analysis of the IR reflectivity indicates that the optical conductivity evolves significantly, reaching the highest conductivity at ~ 40 GPa. Our results highlight pressure's dramatic role in tuning the structural and electronic state of silver chacogenides.

0 citations0 reviewsNo code linkedOpen access
preprint2016arXiv

Scintillation Properties and Electronic Structure of Intrinsic and Extrinsic Mixed Elpasolites Cs2NaRBr3I3 (R = La, Y)

Scintillators attract wide research interest for their distinct applications in radiation detection. Elpasolite halides are among the most promising scintillators due to their high structural symmetry and good scintillation performance. Better understanding of their underlying scintillation mechanism opens up new possibilities in scintillator development. In this work, we employ a variety of experimental techniques to study the two mixed-anion elpasolites Cs2NaRBr3I3 (R = La, Y). The emission of intrinsic Cs2NaRBr3I3 with a light yield ranging from 20,000 to 40,000 ph/MeV is dominant by self-trapped exciton emission. Partial substitution of R with Ce introduces a competing emission, the Ce3+ 5d to 4f radiative transition. Ab initio calculations were performed to investigate the electronic structures as well as the binding energies of polarons in Cs2NaRBr6. The calculated large self-trapped exciton binding energies are consistent with the observed high light yield due to self-trapped exciton emission. The unique electronic structure of halide elpasolites as calculated enhances the STE stability and the STE emission. The highly tunable scintillation properties of mixed-anion elpasolites underscore the role of their complex scintillation mechanism. Our study provides guidance for the design of new elpasolites scintillators with exceptional energy resolution and light yield desirable for applications.

0 citations0 reviewsNo code linkedOpen access
preprint2015arXiv

Fano resonance assisting plasmonic circular dichroism from nanorice heterodimers for extrinsic chirality

In this work, the circular dichroisms (CD) of nanorice heterodimers consisting of two parallel arranged nanorices with the same size but different materials are investigated theoretically. Symmetry-breaking is introduced by using different materials and oblique incidence to achieve strong CD at the vicinity of Fano resonance peaks. We demonstrate that all Au-Ag heterodimers exhibit multipolar Fano resonances and strong CD effect. A simple quantitative analysis shows that the structure with larger Fano asymmetry factor has stronger CD. The intensity and peak positions of the CD effect can be flexibly tuned in a large range by changing particle size, shape, the inter-particle distance and surroundings. Furthermore, CD spectra exhibit high sensitivity to ambient medium in visible and near infrared regions. Our results here are beneficial for the design and application of high sensitive CD sensors and other related fields.

0 citations0 reviewsNo code linkedOpen access
preprint2010arXiv

Quantum information processing architecture with endohedral fullerenes in a carbon nanotube

A potential quantum information processor is proposed using a fullerene peapod, i.e., an array of the endohedral fullerenes 15N@C60 or 31P@C60 contained in a single walled carbon nanotube (SWCNT). The qubits are encoded in the nuclear spins of the doped atoms, while the electronic spins are used for initialization and readout, as well as for two-qubit operations.

0 citations0 reviewsNo code linkedOpen access