MokA

Search documents
Why Altria Stock Is Sinking Today
The Motley Fool· 2025-07-09 19:19
Core Viewpoint - Altria Group's stock is experiencing a significant decline following a negative rating from Jefferies, despite a generally positive market trend [1][3]. Group 1: Stock Performance - Altria's stock price fell by 4.1% as of 2:30 p.m. ET, with a maximum decline of 4.5% earlier in the trading session [1]. - Despite the current pullback, Altria's stock has increased approximately 24% over the past year [1]. Group 2: Analyst Coverage - Jefferies initiated coverage on Altria with an underperform rating and set a one-year price target of $50 per share, indicating a potential downside of about 12.5% [3]. - The lead analyst, Edward Mundy, believes that Altria's valuation has become stretched following its recent stock price increase [3]. Group 3: Market Trends - Altria is facing a potential secular decline in smokable tobacco products, with consistent declines in unit sales for cigarettes in recent years [4]. - The company has been able to offset some declines in cigarette unit volumes through price increases [4]. Group 4: Future Outlook - Altria is focusing on increasing demand for smoke-free products and has seen some recent successes in this area [5]. - The company's current valuation stands at approximately 10.6 times this year's expected earnings, with a dividend yield of around 7.2% [5].
2025外贸企业HR系统选型指南:Moka全球化功能解析
Sou Hu Cai Jing· 2025-07-09 07:51
全球化招聘管理:精准触达国际人才 外贸企业拓展全球市场,对熟悉不同国家文化、市场规则及具备多语言能力的人才求贤若渴。传统招聘方式受地域限制,难以高效吸引优 质国际人才,招聘渠道分散也增加了管理难度。2025 年,HR 系统需具备强大的全球化招聘功能,整合全球资源,精准匹配岗位需求。 Moka 薪酬模块涵盖 120 + 国家薪资计算规则,支持多币种结算,能根据各国税法实时更新,自动生成合规报表。某外贸企业使用 Moka 后,海外薪酬核算时间从 5 天缩短至 1 天,合规问题减少 90%,有效降低了运营风险,保障了企业全球业务稳定开展。 外贸企业团队分散于世界各地,不同国家和地区的法规政策、工作习惯差异大,人事管理面临数据分散、流程混乱等难题。传统系统难以 满足多地域协同管理需求,导致信息滞后、管理效率低下。2025 年,HR 系统需实现全球人事数据统一管理,支持多语言交互与本地化流 程适配。 Moka 人事模块可整合全球分支机构员工信息,支持 20 + 语言自由切换,打破语言障碍。系统自动适配各国考勤规则、假期制度,实时同 步员工信息变更,确保数据准确一致。某跨国外贸集团运用 Moka 后,跨地域人事审批效率提 ...
充分激发模态协作,MokA量身打造MLLM微调新范式
机器之心· 2025-06-29 02:21
本文第一作者卫雅珂为中国人民大学四年级博士生,主要研究方向为多模态学习机制、多模态大模型等,师从胡迪副教授。作者来自于中国人民大学和上海人工 智能实验室。 近年来,多模态大模型(MLLMs)已经在视觉语言、音频语言等任务上取得了巨大进展。然而 ,当在多模态下游任务进行微调时,当前主流的多模态微调方法大 多直接沿用了在纯文本大语言模型(LLMs)上发展出的微调策略,比如 LoRA。但这种「照搬」 策略,真的适用于多模态模型吗? 多基座、多场景下均实现性能提升 论文标题:MokA: Multimodal Low-Rank Adaptation for MLLMs 论文链接:https://arxiv.org/abs/2506.05191 项目主页:https://gewu-lab.github.io/MokA 来自中国人民大学高瓴人工智能学院 GeWu-Lab 实验室、上海人工智能实验室的研究团队在最新论文中给出了一种全新的思考方式。他们指出:当下 MLLMs 微调 方案大多简单的将单模态策略迁移至多模态场景,未结合多模态学习特性进行深入思考。事实 上,在多模态场景中,单模态信息的独立建模(Unimodal Ad ...
Mogo(MOGO) - 2024 Q4 - Earnings Call Transcript
2025-03-20 18:47
Mogo Inc. (NASDAQ:MOGO) Q4 2024 Earnings Conference Call March 20, 2025 10:30 AM ET Company Participants Craig Armitage ??? Investor Relations Dave Feller ??? Chief Executive Officer Greg Feller ??? President and Chief Financial Officer Conference Call Participants Scott Buck ??? H.C. Wainwright Operator Good morning, ladies and gentlemen, and welcome to Mogo Inc. Q4 2024 Financial Results Conference Call. At this time, all lines are in listen-only mode. Following the presentation, we will conduct a questio ...