Hasil untuk "Manners and customs (General)"

Yuchuan Ye, Ming Ding, Youjia Chen et al.

In this article, we explore federated customization of large models and highlight the key challenges it poses within the federated learning framework. We review several popular large model customization techniques, including full fine-tuning, efficient fine-tuning, prompt engineering, prefix-tuning, knowledge distillation, and retrieval-augmented generation. Then, we discuss how these techniques can be implemented within the federated learning framework. Moreover, we conduct experiments on federated prefix-tuning, which, to the best of our knowledge, is the first trial to apply prefix-tuning in the federated learning setting. The conducted experiments validate its feasibility with performance close to centralized approaches. Further comparison with three other federated customization methods demonstrated its competitive performance, satisfactory efficiency, and consistent robustness.

Samuel Koovely

We introduce a queueing-based sampling framework for continuous-time temporal networks. We focus on a Markovian parametrization in which link start times follow a homogeneous Poisson process and link durations are exponentially distributed. We derive stochastic properties of the resulting link streams and exploit them to generate synthetic temporal networks with controllable smoothness and prescribed event patterns, relevant for the validation and interpretation of methods for community, scale, change-point, and periodicity detection. By coupling this temporal mechanism with block-structured endpoint distributions, we obtain a continuous-time analogue of stochastic block models. We also discuss extensions of the framework, including discrete-time and instantaneous-contact limits.

Jie Zhu, Huaixia Dou, Junhui Li et al.

Effective customer support requires not only accurate problem solving but also structured and empathetic communication aligned with professional standards. However, existing dialogue datasets often lack strategic guidance, and real-world service data is difficult to access and annotate. To address this, we introduce the task of Customer Support Conversation (CSC), aimed at training customer service agents to respond using well-defined support strategies. We propose a structured CSC framework grounded in COPC guidelines, defining five conversational stages and twelve strategies to guide high-quality interactions. Based on this, we construct CSConv, an evaluation dataset of 1,855 real-world customer-agent conversations rewritten using LLMs to reflect deliberate strategy use, and annotated accordingly. Additionally, we develop a role-playing approach that simulates strategy-rich conversations using LLM-powered roles aligned with the CSC framework, resulting in the training dataset RoleCS. Experiments show that fine-tuning strong LLMs on RoleCS significantly improves their ability to generate high-quality, strategy-aligned responses on CSConv. Human evaluations further confirm gains in problem resolution. All code and data will be made publicly available at https://github.com/aliyun/qwen-dianjin.

Hanchen Su, Wei Luo, Wei Han et al.

We propose a practical approach by integrating Large Language Models (LLMs) with a framework designed to navigate the complexities of Airbnb customer support operations. In this paper, our methodology employs a novel reformatting technique, the Intent, Context, and Action (ICA) format, which transforms policies and workflows into a structure more comprehensible to LLMs. Additionally, we develop a synthetic data generation strategy to create training data with minimal human intervention, enabling cost-effective fine-tuning of our model. Our internal experiments (not applied to Airbnb products) demonstrate that our approach of restructuring workflows and fine-tuning LLMs with synthetic data significantly enhances their performance, setting a new benchmark for their application in customer support. Our solution is not only cost-effective but also improves customer support, as evidenced by both accuracy and manual processing time evaluation metrics.

Yanwei Sun, Zhe Liu, Chiwei Yan

Motivated by applications such as urban traffic control and make-to-order systems, we study a fluid model of a single-server, on-off system that can accommodate multiple queues. The server visits each queue in order: when a queue is served, it is "on", and when the server is serving another queue or transitioning between queues, it is "off". Customers arrive over time, observe the state of the system, and decide whether to join. We consider two regimes for the formation of the on and off durations. In the exogenous setting, each queue's on and off durations are predetermined. We explicitly characterize the equilibrium outcome in closed form and give a compact linear program to compute the optimal on-off durations that maximizes total reward collected from serving customers. In the endogenous setting, the durations depend on customers' joining decisions under an exhaustive service policy where the server never leaves a non-empty queue. We show that an optimal policy in this case extends service beyond the first clearance for at most one queue. Using this property, we introduce a closed-form procedure that computes an optimal policy in no more than 2n steps for a system with n queues.

Xiaoqun Liu, Jiacheng Liang, Luoxi Tang et al.

Large language models (LLMs) are widely adapted for downstream applications through fine-tuning, a process named customization. However, recent studies have identified a vulnerability during this process, where malicious samples can compromise the robustness of LLMs and amplify harmful behaviors-an attack commonly referred to as jailbreaking. To address this challenge, we propose an adaptive data curation approach allowing any text to be curated to enhance its effectiveness in counteracting harmful samples during customization. To avoid the need for additional defensive modules, we further introduce a comprehensive mitigation framework spanning the lifecycle of the customization process: before customization to immunize LLMs against future jailbreak attempts, during customization to neutralize risks, and after customization to restore compromised models. Experimental results demonstrate a significant reduction in jailbreaking effects, achieving up to a 100% success rate in generating safe responses. By combining adaptive data curation with lifecycle-based mitigation strategies, this work represents a solid step forward in mitigating jailbreaking risks and ensuring the secure adaptation of LLMs.

Geonhui Jang, Jin-Hwa Kim, Yong-Hyun Park et al.

Text-to-image (T2I) models can effectively capture the content or style of reference images to perform high-quality customization. A representative technique for this is fine-tuning using low-rank adaptations (LoRA), which enables efficient model customization with reference images. However, fine-tuning with a limited number of reference images often leads to overfitting, resulting in issues such as prompt misalignment or content leakage. These issues prevent the model from accurately following the input prompt or generating undesired objects during inference. To address this problem, we examine the text embeddings that guide the diffusion model during inference. This study decomposes the text embedding matrix and conducts a component analysis to understand the embedding space geometry and identify the cause of overfitting. Based on this, we propose DECOR, which projects text embeddings onto a vector space orthogonal to undesired token vectors, thereby reducing the influence of unwanted semantics in the text embeddings. Experimental results demonstrate that DECOR outperforms state-of-the-art customization models and achieves Pareto frontier performance across text and visual alignment evaluation metrics. Furthermore, it generates images more faithful to the input prompts, showcasing its effectiveness in addressing overfitting and enhancing text-to-image customization.

Tianchu Guo, Pengyu Li, Biao Wang et al.

Recently customized generation has significant potential, which uses as few as 3-5 user-provided images to train a model to synthesize new images of a specified subject. Though subsequent applications enhance the flexibility and diversity of customized generation, fine-grained control over the given subject acting like the person's pose is still lack of study. In this paper, we propose a PersonificationNet, which can control the specified subject such as a cartoon character or plush toy to act the same pose as a given referenced person's image. It contains a customized branch, a pose condition branch and a structure alignment module. Specifically, first, the customized branch mimics specified subject appearance. Second, the pose condition branch transfers the body structure information from the human to variant instances. Last, the structure alignment module bridges the structure gap between human and specified subject in the inference stage. Experimental results show our proposed PersonificationNet outperforms the state-of-the-art methods.

Andrew Estornell, Stylianos Loukas Vasileiou, William Yeoh et al.

In today's competitive financial landscape, understanding and anticipating customer goals is crucial for institutions to deliver a personalized and optimized user experience. This has given rise to the problem of accurately predicting customer goals and actions. Focusing on that problem, we use historical customer traces generated by a realistic simulator and present two simple models for predicting customer goals and future actions -- an LSTM model and an LSTM model enhanced with state-space graph embeddings. Our results demonstrate the effectiveness of these models when it comes to predicting customer goals and actions.

Luozhou Wang, Ziyang Mai, Guibao Shen et al.

In this work, we present a novel approach for motion customization in video generation, addressing the widespread gap in the exploration of motion representation within video generative models. Recognizing the unique challenges posed by the spatiotemporal nature of video, our method introduces Motion Embeddings, a set of explicit, temporally coherent embeddings derived from a given video. These embeddings are designed to integrate seamlessly with the temporal transformer modules of video diffusion models, modulating self-attention computations across frames without compromising spatial integrity. Our approach provides a compact and efficient solution to motion representation, utilizing two types of embeddings: a Motion Query-Key Embedding to modulate the temporal attention map and a Motion Value Embedding to modulate the attention values. Additionally, we introduce an inference strategy that excludes spatial dimensions from the Motion Query-Key Embedding and applies a differential operation to the Motion Value Embedding, both designed to debias appearance and ensure the embeddings focus solely on motion. Our contributions include the introduction of a tailored motion embedding for customization tasks and a demonstration of the practical advantages and effectiveness of our method through extensive experiments.

Chenxi Liu, Gan Sun, Wenqi Liang et al.

Pre-trained large text-to-image (T2I) models with an appropriate text prompt has attracted growing interests in customized images generation field. However, catastrophic forgetting issue make it hard to continually synthesize new user-provided styles while retaining the satisfying results amongst learned styles. In this paper, we propose MuseumMaker, a method that enables the synthesis of images by following a set of customized styles in a never-end manner, and gradually accumulate these creative artistic works as a Museum. When facing with a new customization style, we develop a style distillation loss module to extract and learn the styles of the training data for new image generation. It can minimize the learning biases caused by content of new training images, and address the catastrophic overfitting issue induced by few-shot images. To deal with catastrophic forgetting amongst past learned styles, we devise a dual regularization for shared-LoRA module to optimize the direction of model update, which could regularize the diffusion model from both weight and feature aspects, respectively. Meanwhile, to further preserve historical knowledge from past styles and address the limited representability of LoRA, we consider a task-wise token learning module where a unique token embedding is learned to denote a new style. As any new user-provided style come, our MuseumMaker can capture the nuances of the new styles while maintaining the details of learned styles. Experimental results on diverse style datasets validate the effectiveness of our proposed MuseumMaker method, showcasing its robustness and versatility across various scenarios.

Zhexiao Xiong, Wei Xiong, Jing Shi et al.

Recent approaches in text-to-image customization have primarily focused on preserving the identity of the input subject, but often fail to control the spatial location and size of objects. We introduce GroundingBooth, which achieves zero-shot, instance-level spatial grounding on both foreground subjects and background objects in the text-to-image customization task. Our proposed grounding module and subject-grounded cross-attention layer enable the creation of personalized images with accurate layout alignment, identity preservation, and strong text-image coherence. In addition, our model seamlessly supports personalization with multiple subjects. Our model shows strong results in both layout-guided image synthesis and text-to-image customization tasks. The project page is available at https://groundingbooth.github.io.

Karin Noemi Rühle Indart, Marcelo Caetano de Sousa

Num contexto de reconstrução do sistema educativo, como aquele que vem caracterizando a situação em Timor-Leste, a Formação de Professores do Ensino Básico em Exercício defronta-se com o problema do recrutamento dos professores de todos os níveis do ensino definido por critérios rigorosos conforme a Lei de Bases da Educação. Por essa razão, o investimento maior do Estado tem sido na formação de professores que já estão no sistema. Assim, o objetivo desta pesquisa é analisar os processos e negociações para a mudança das Políticas de Formação de Professores do Ensino Básico em Exercício, mais especificamente do Curso de Bacharelato dos Professores em Regime de Contrato, do 1º e 2º Ciclos do Ensino Básico, em todos os municípios de Timor-Leste. A abordagem de pesquisa utilizada é a qualitativa-descritiva e usa como fonte de dados entrevistas com os gestores nacionais destas políticas de formação. A análise das entrevistas mostra que a iniciativa de criação e responsabilidade de gestão de um novo programa foi do Ministério da Educação, Juventude e Desporto, porém, o INFORDEPE e a UNTL estiveram envolvidos em todas as fases de discussão. Este programa foi elaborado para resolver uma situação criada no período de emergência – a contratação de professores voluntários sem formação inicial. Concluímos que o período alongado e a intensa disposição de encontros e conversas entre os gestores das instituições demonstram que o programa foi de fato pensado e planejado, diferente de ações emergenciais realizadas anteriormente. Porém, o modelo e o objetivo principal da formação não diferem realmente das políticas anteriores.

Maitreyee Marathe, Line A. Roald

For a large (and recently increasing) number of households, affordability is a major hurdle in accessing sufficient electricity and avoiding service disconnections. For such households, in-home energy rationing, i.e. the need to actively prioritize how to use a limited amount of electricity, is an everyday reality. In this paper, we consider a particularly vulnerable group of customers, namely prepaid electricity customers, who are required to pay for their electricity a-priori. With this group of customers in mind, we propose an optimization-based energy management framework to effectively use a limited budget and avoid the disruptions and fees associated with disconnections. The framework considers forecasts of future use and knowledge of appliance power ratings to help customers prioritize and limit use of low-priority loads, with the goal of extending access to their critical loads. Importantly, the proposed management system has minimal requirements in terms of in-home hardware and remote communication, lending itself well to adoption across different regions and utility programs. Our case study demonstrates that by considering both current and future electricity consumption and more effectively managing access to low-priority loads, the proposed framework increases the value provided to customers and avoids disconnections.

Marie Claire De Mattia

Dentro da constelação literária dos últimos anos, um lugar de visibilidade especial cabe ao romance O alegre canto da perdiz (2016), de autoria da moçambicana Paulina Chiziane. Esta obra articulada e intensa encena a História moçambicana desde a colonização até à época da independência – e fá-lo contando as vivências atribuladas das quatro gerações que compõem uma família moçambicana. Este artigo se debruça sobre a representação da violência na obra, estudando tanto as violências de género como as de raça. Analisamos aqui como e em que medida a prática constante da violência (física, psicológica e epistémica) durante o regime colonial afeta o processo de busca e de construção identitária do ponto de vista nacional como também individual, tanto para os homens como para as mulheres. Além disso, este trabalho reflete sobre como para estas últimas as brutalidades sistemáticas da dominação portuguesa acabam por desestabilizar as categorias ontológicas quer da gestão da própria sexualidade, quer da maternidade.

Xin Yuan, Yongbing Feng, Mingming Ye et al.

Custom voice is to construct a personal speech synthesis system by adapting the source speech synthesis model to the target model through the target few recordings. The solution to constructing a custom voice is to combine an adaptive acoustic model with a robust vocoder. However, training a robust vocoder usually requires a multi-speaker dataset, which should include various age groups and various timbres, so that the trained vocoder can be used for unseen speakers. Collecting such a multi-speaker dataset is difficult, and the dataset distribution always has a mismatch with the distribution of the target speaker dataset. This paper proposes an adaptive vocoder for custom voice from another novel perspective to solve the above problems. The adaptive vocoder mainly uses a cross-domain consistency loss to solve the overfitting problem encountered by the GAN-based neural vocoder in the transfer learning of few-shot scenes. We construct two adaptive vocoders, AdaMelGAN and AdaHiFi-GAN. First, We pre-train the source vocoder model on AISHELL3 and CSMSC datasets, respectively. Then, fine-tune it on the internal dataset VXI-children with few adaptation data. The empirical results show that a high-quality custom voice system can be built by combining a adaptive acoustic model with a adaptive vocoder.

Administrador Veredas

Ana Carolina Lopes Costa

Este artigo tem por objetivo realizar uma leitura, guiada pelo princípio plagiotrópico, da tradução das duas cenas finais do Fausto II, de Goethe, efetuada por Haroldo de Campos. O termo plagiotropia, oriundo do campo biológico, foi aproveitado pelo crítico, tradutor e poeta paulista para explicar o desenvolvimento da tradição literária, lançando sobre essa uma noção de plasticidade: a força motriz que permite o desenvolvimento e a conjugação das poéticas ao longo do tempo é também transversal (Campos, 2005). Aqui, do mesmo modo, o estudo dessa união de poéticas passa necessariamente pelo conceito haroldiano de transcriação: a tradução como criação e como crítica. Nesta esteira, focalizando a intertextualidade de Goethe, que se utiliza do tom dos coveiros hamletianos na cena "Enterramento", Campos, pautado pela noção de tradução como crítica, transcriará a cena através do diálogo com João Cabral de Melo Neto. O coro dos lêmures, em português, está imbricado pelos versos de Morte e vida severina intensificando, assim, o processo de ramificação oblíqua. Para efeito de compressão desse gesto tradutório, refletiremos sobre o que chamamos de dicção da pá, a concretude cabralina, objetivando compreender como Haroldo de Campos o aproveita em sua tradução. Em sua prática transluciferina, Campos adota a fissura da forma e do conteúdo para usurpar, pelas frestas do discurso, o trono do texto original.

Juliana Santini

Resenha: Living (il)legalities in Brazil, com organização de Sara Brandellero, Derek Pardue e Georg Wink.

Canqi Yao, Shibo Chen, Zaiyue Yang

In this paper, we study the cooperative operation problem between the fleet operator and incentive-aware customers in an on-demand delivery system. Specifically, the fleet operator offers discounts on transportation costs in exchange of the delivery time flexibility of customers. In order to capture the interaction between the fleet operator and customers, a novel bi-level optimization framework is proposed. By exploiting the strong duality, and the KKT optimality condition of customer optimization problems, we can reformulate the bi-level optimization problem as a mixed integer nonlinear programming problem. Considering the inherent difficulties of MINLP, a computationally efficient algorithm, which combines the merits of Lagrangian dual decomposition and Benders decomposition, is devised to solve the resulting MINLP problem in a distributed manner. Finally, extensive numerical experiments demonstrate that the proposed cooperation scheme can decrease the delivery fees for the customers, and reduce the operation cost of the fleet operator at the same time, thus leading to a win-win situation for both sides.