arXiv subjects

Hongyu Xiong

Publications and source records attributed to Hongyu Xiong.

14 recordsLinked to original sources

arXiv2026

When Rules Fall Short: Agent-Driven Discovery of Emerging Content Issues in Short Video Platforms

Trends on short-video platforms evolve at a rapid pace, with new content issues emerging every day that fall outside the coverage of existing annotation policies. However, traditional human-driven discovery of emerging issues is too slow, which leads to delayed updates of annotation policies and poses a major challenge for effective content governance. In this work, we propose an automatic issue discovery method based on multimodal LLM agents. Our approach automatically recalls short videos containing potential new issues and applies a two-stage clustering strategy to group them, with each cluster corresponding to a newly discovered issue. The agent then generates updated annotation policies from these clusters, thereby extending coverage to these emerging issues. Our agent has been deployed in the real system. Both offline and online experiments demonstrate that this agent-based method significantly improves the effectiveness of emerging-issue discovery (with an F1 score improvement of over 20%) and enhances the performance of subsequent issue governance (reducing the view count of problematic videos by approximately 15%). More importantly, compared to manual issue discovery, it greatly reduces time costs and substantially accelerates the iteration of annotation policies.

cs.CV↗

arXiv2025

Reasoning-Enhanced Domain-Adaptive Pretraining of Multimodal Large Language Models for Short Video Content Governance

Short video platforms are evolving rapidly, making the identification of inappropriate content increasingly critical. Existing approaches typically train separate and small classification models for each type of issue, which requires extensive human-labeled data and lacks cross-issue generalization. We propose a reasoning-enhanced multimodal large language model (MLLM) pretraining paradigm for unified inappropriate content detection. To address the distribution gap between short video content and the original pretraining data of MLLMs, as well as the complex issue definitions, we introduce three targeted pretraining tasks: (1) \textit{Caption}, to enhance the MLLM's perception of video details; (2) \textit{Visual Question Answering (VQA)}, to deepen the MLLM's understanding of issue definitions and annotation guidelines; (3) \textit{Chain-of-Thought (CoT)}, to enhance the MLLM's reasoning capability. Experimental results show that our pretraining approach significantly improves the MLLM's performance in both zero-shot and supervised fine-tuning (SFT) settings. In addition, our pretrained model demonstrates strong generalization capabilities to emergent, previously unseen issues.

cs.CV↗

arXiv2025

Filter-And-Refine: A MLLM Based Cascade System for Industrial-Scale Video Content Moderation

Effective content moderation is essential for video platforms to safeguard user experience and uphold community standards. While traditional video classification models effectively handle well-defined moderation tasks, they struggle with complicated scenarios such as implicit harmful content and contextual ambiguity. Multimodal large language models (MLLMs) offer a promising solution to these limitations with their superior cross-modal reasoning and contextual understanding. However, two key challenges hinder their industrial adoption. First, the high computational cost of MLLMs makes full-scale deployment impractical. Second, adapting generative models for discriminative classification remains an open research problem. In this paper, we first introduce an efficient method to transform a generative MLLM into a multimodal classifier using minimal discriminative training data. To enable industry-scale deployment, we then propose a router-ranking cascade system that integrates MLLMs with a lightweight router model. Offline experiments demonstrate that our MLLM-based approach improves F1 score by 66.50% over traditional classifiers while requiring only 2% of the fine-tuning data. Online evaluations show that our system increases automatic content moderation volume by 41%, while the cascading deployment reduces computational cost to only 1.5% of direct full-scale deployment.

cs.LG↗

arXiv2025

COEF-VQ: Cost-Efficient Video Quality Understanding through a Cascaded Multimodal LLM Framework

Recently, with the emergence of recent Multimodal Large Language Model (MLLM) technology, it has become possible to exploit its video understanding capability on different classification tasks. In practice, we face the difficulty of huge requirements for GPU resource if we need to deploy MLLMs online. In this paper, we propose COEF-VQ, a novel cascaded MLLM framework designed to enhance video quality understanding on the short-video platform while optimizing computational efficiency. Our approach integrates an entropy-based pre-filtering stage, where a lightweight model assesses uncertainty and selectively filters cases before passing them to the more computationally intensive MLLM for final evaluation. By prioritizing high-uncertainty samples for deeper analysis, our framework significantly reduces GPU usage while maintaining the strong classification performance of a full MLLM deployment. To demonstrate the effectiveness of COEF-VQ, we deploy this new framework onto the video management platform (VMP) at the short-video platform, and perform a series of detailed experiments on two in-house tasks related to video quality understanding. We show that COEF-VQ leads to substantial performance gains from the offline evaluation in these two tasks and effectively enhances platform safety with limit resource consumption, significantly reducing inappropriate content video view rate by 9.9% in a online A/B test without affecting engagement. Post-launch monitoring confirmed sustained improvements, validating its real-world impact.

cs.CV↗

arXiv2025

Audio-Enhanced Vision-Language Modeling with Latent Space Broadening for High Quality Data Expansion

Transformer-based multimodal models are widely used in industrial-scale recommendation, search, and advertising systems for content understanding and relevance ranking. Enhancing labeled training data quality and cross-modal fusion significantly improves model performance, influencing key metrics such as quality view rates and ad revenue. High-quality annotations are crucial for advancing content modeling, yet traditional statistical-based active learning (AL) methods face limitations: they struggle to detect overconfident misclassifications and are less effective in distinguishing semantically similar items in deep neural networks. Additionally, audio information plays an increasing role, especially in short-video platforms, yet most pre-trained multimodal architectures primarily focus on text and images. While training from scratch across all three modalities is possible, it sacrifices the benefits of leveraging existing pre-trained visual-language (VL) and audio models. To address these challenges, we propose kNN-based Latent Space Broadening (LSB) to enhance AL efficiency and Vision-Language Modeling with Audio Enhancement (VLMAE), a mid-fusion approach integrating audio into VL models. This system deployed in production systems, leading to significant business gains.

cs.MM↗

arXiv2025

USM: Unbiased Survey Modeling for Limiting Negative User Experiences in Recommendation Systems

Reducing negative user experiences is essential for the success of recommendation platforms. Exposing users to inappropriate content could not only adversely affect users' psychological well-beings, but also potentially drive users away from the platform, sabotaging the platform's long-term success. However, recommendation algorithms tend to weigh more heavily on positive feedback signals due to the scarcity of negative ones, which may result in the neglect of valuable negative user feedback. In this paper, we propose an approach aimed at limiting negative user experiences. Our method primarily relies on distributing in-feed surveys to the users, modeling the users' feedback collected from the survey, and integrating the model predictions into the recommendation system. We further enhance the baseline survey model by integrating the Learning Hidden Unit Contributions module and the Squeeze-and-Excitation module. In addition, we strive to resolve the problem of response Bias by applying a survey-submit model; The A/B testing results indicate a reduction in survey sexual rate and survey inappropriate rate, ranging from -1.44\% to -3.9\%. Additionally, we compared our methods against an online baseline that does not incorporate our approach. The results indicate that our approach significantly reduces the report rate and dislike rate by 1\% to 2.27\% compared to the baseline, confirming the effectiveness of our methods in enhancing user experience. After we launched the survey model based our approach on our platform, the model is able to bring reductions of 1.75\%, 2.57\%, 2.06\% on reports, dislikes, survey inappropriate rate, respectively.

cs.IR↗

arXiv2024

IPS: In-Prompt Process Supervision for Short Video Content Moderation

Multimodal large language models (MLLMs) are effective at capturing the semantics of short video content; however, they often fail to attend to the policy-specific details required for reliable content moderation. To address this limitation, we introduce IPS, a novel framework that integrates In-prompt Process Supervision into MLLMs by introducing sequential reasoning over ancillary questions during fine-tuning. IPS consistently outperforms baseline MLLMs on public and proprietary benchmarks. Moreover, replacing human-annotated ancillary labels with MLLM-generated ones results in only marginal performance degradation, demonstrating robustness to noisy supervision and strong scalability with model-generated annotations. These findings establish IPS as a scalable and effective solution for complex multimodal classification in large-scale industrial settings.

cs.CL↗

arXiv2022

Evidences for the exciton gas phase and its condensation in monolayer 1T-ZrTe2

The excitonic insulator (EI) is a Bose-Einstein condensation (BEC) of excitons bound by electron-hole interaction in a solid, which could support high-temperature BEC transition. The material realization of EI has been elusive, which is further challenged by the difficulty of distinguishing it from a conventional charge density wave (CDW) state. In the BEC limit, the pre-condensation exciton gas phase is a hallmark to distinguish EI from conventional CDW, yet direct experimental evidence has been lacking. Here we report a distinct correlated phase beyond the $2\times2$ CDW ground state emerging in epitaxially grown monolayer 1T-ZrTe2 and its investigation by angle-resolved photoemission spectroscopy (ARPES) and scanning tunneling microscopy (STM). The results show novel band- and energy-dependent folding behavior in a two-step process, evidenced by an exciton gas phase prior to its condensation into the final CDW state. The excellent agreement between experiments and theoretical predictions on the recovery of the pristine band structure by carrier-density-dependent suppression of the CDW state further corroborates the monolayer 1T-ZrTe2 as an EI. Our findings provide a versatile two-dimensional platform that allows tuning of the excitonic effect.

cond-mat.mtrl-sci↗

arXiv2019

Visualizing Exotic Orbital Texture in the Single-Layer Mott Insulator 1T-TaSe2

Mott insulating behavior is induced by strong electron correlation and can lead to exotic states of matter such as unconventional superconductivity and quantum spin liquids. Recent advances in van der Waals material synthesis enable the exploration of novel Mott systems in the two-dimensional limit. Here we report characterization of the local electronic properties of single- and few-layer 1T-TaSe2 via spatial- and momentum-resolved spectroscopy involving scanning tunneling microscopy and angle-resolved photoemission. Our combined experimental and theoretical study indicates that electron correlation induces a robust Mott insulator state in single-layer 1T-TaSe2 that is accompanied by novel orbital texture. Inclusion of interlayer coupling weakens the insulating phase in 1T-TaSe2, as seen by strong reduction of its energy gap and quenching of its correlation-driven orbital texture in bilayer and trilayer 1T-TaSe2. Our results establish single-layer 1T-TaSe2 as a useful new platform for investigating strong correlation physics in two dimensions.

cond-mat.str-el↗

arXiv2018

Efficient Super Resolution Using Binarized Neural Network

Deep convolutional neural networks (DCNNs) have recently demonstrated high-quality results in single-image super-resolution (SR). DCNNs often suffer from over-parametrization and large amounts of redundancy, which results in inefficient inference and high memory usage, preventing massive applications on mobile devices. As a way to significantly reduce model size and computation time, binarized neural network has only been shown to excel on semantic-level tasks such as image classification and recognition. However, little effort of network quantization has been spent on image enhancement tasks like SR, as network quantization is usually assumed to sacrifice pixel-level accuracy. In this work, we explore an network-binarization approach for SR tasks without sacrificing much reconstruction accuracy. To achieve this, we binarize the convolutional filters in only residual blocks, and adopt a learnable weight for each binary filter. We evaluate this idea on several state-of-the-art DCNN-based architectures, and show that binarized SR networks achieve comparable qualitative and quantitative results as their real-weight counterparts. Moreover, the proposed binarized strategy could help reduce model size by 80% when applying on SRResNet, and could potentially speed up inference by 5 times.

cs.CV↗

arXiv2018

Transferable Natural Language Interface to Structured Queries aided by Adversarial Generation

A natural language interface (NLI) to structured query is intriguing due to its wide industrial applications and high economical values. In this work, we tackle the problem of domain adaptation for NLI with limited data on target domain. Two important approaches are considered: (a) effective general-knowledge-learning on source domain semantic parsing, and (b) data augmentation on target domain. We present a Structured Query Inference Network (SQIN) to enhance learning for domain adaptation, by separating schema information from NL and decoding SQL in a more structural-aware manner; we also propose a GAN-based augmentation technique (AugmentGAN) to mitigate the issue of lacking target domain data. We report solid results on GeoQuery, Overnight, and WikiSQL to demonstrate state-of-the-art performances for both in-domain and domain-transfer tasks.

cs.CL↗

arXiv2018

A Transfer-Learnable Natural Language Interface for Databases

Relational database management systems (RDBMSs) are powerful because they are able to optimize and answer queries against any relational database. A natural language interface (NLI) for a database, on the other hand, is tailored to support that specific database. In this work, we introduce a general purpose transfer-learnable NLI with the goal of learning one model that can be used as NLI for any relational database. We adopt the data management principle of separating data and its schema, but with the additional support for the idiosyncrasy and complexity of natural languages. Specifically, we introduce an automatic annotation mechanism that separates the schema and the data, where the schema also covers knowledge about natural language. Furthermore, we propose a customized sequence model that translates annotated natural language queries to SQL statements. We show in experiments that our approach outperforms previous NLI methods on the WikiSQL dataset and the model we learned can be applied to another benchmark dataset OVERNIGHT without retraining.

cs.AI↗

arXiv2018

Band resolved imaging of photocurrent in a topological insulator

We study the microscopic origins of photocurrent generation in the topological insulator Bi$_2$Se$_3$ via time- and angle-resolved photoemission spectroscopy. We image the unoccupied band structure as it evolves following a circularly polarized optical excitation and observe an asymmetric electron population in momentum space, which is the spectroscopic signature of a photocurrent. By analyzing the rise times of the population we identify which occupied and unoccupied electronic states are coupled by the optical excitation. We conclude that photocurrents can only be excited via resonant optical transitions coupling to spin-orbital textured states. Our work provides a microscopic understanding of how to control photocurrents in systems with spin-orbit coupling and broken inversion symmetry.

cond-mat.mtrl-sci↗

arXiv2017

Anomalous Hall Effect in ZrTe5

ZrTe$_5$ has been of recent interest as a potential Dirac/Weyl semimetal material. Here, we report the results of experiments performed via in-situ 3D double-axis rotation to extract the full $4π$ solid angular dependence of the transport properties. A clear anomalous Hall effect (AHE) was detected for every sample, with no magnetic ordering observed in the system to the experimental sensitivity of torque magnetometry. Interestingly, the AHE takes large values when the magnetic field is rotated in-plane, with the values vanishing above $\sim 60$ K where the negative longitudinal magnetoresistance (LMR) also disappears. This suggests a close relation in their origins, which we attribute to Berry curvature generated by the Weyl nodes.

cond-mat.mes-hall↗