广场

全村希望,千问 4 Max 正式发布,2.4 万亿参数模型「进厂打工」

作者|桦林舞王 编辑|靖宇 底座架构重写,单任务成本暴降近六成,阿里把旗舰模型拉进了真实排班表。 终于,一款国产旗舰模型现身了。 11 月 16 日,阿里千问正式发布新一代旗舰模型 Qwen4-Max。就在五天前,它的预览版 qwen4-max-preview 刚在 Token Plan 里低调露过面,今天正式版直接亮出了完整底牌。 新模型依然维持着 2.4 万亿(2.4T)的主干总参数规模,原生支持文本、图像和视频输入,上下文窗口拉到了 100 万 Token。不过比参数表更显眼的,是官方同时给出的一笔账单:在模拟真实工作场景的评测里,千问 4 Max 跑完单个任务的平均费用, 从上一代的 26.58 元降到了 11.16 元,直接砍掉 58% 。 千问团队这次释放的信号很明确,旗舰大模型单纯拼纸面考分的阶段翻篇了,现在的目标是把模型塞进真实工作流,当一个能扛长活、也算得过账的「新同事」。 01 给 2.4 万亿参数的巨兽「减负」 过去一年,行业里没少见识号称百万上下文的模型。但真把几十万字的项目文档或者一整个代码库丢进去,很多模型到了后半程就开始注意力涣散、输出卡顿,账单还贵得吓人。要让「新同事」在 100 万上下文和长程 Agent 任务里跑得稳、跑得省,光靠外围工程补丁已经不够了,必须动底层架构。 千问 4 Max 这次在底座上做了三处关键调整,核心逻辑都是把算力花在刀刃上。 先看负责记忆的注意力层。如果让模型在长文本里每写一个字,都要回头把前面几十万字逐字重读一遍,算力开销会急剧膨胀。千问换上了混合方案(GDN 加 Qwen Sparse Attention),每 4…

极客公园AI 工具中文

‘Meet the Anthropic Researcher Behind Silicon Valley’s “Consensual Non-Consent” Orgies’

Aaron Sibarium, writing for The Free Beacon: Paul Crowley, a cybersecurity researcher at Anthropic, is listed as a manager of Red Means No LLC, which says that its parties are “designed to be spaces to … ignore the ‘nos’ of others.” Participants can use a safe word, “red,” to end an encounter but give “default consent” to “spanking,” “wrestling,” and “light pain” by entering the orgy, according to Red Means No’s website . More extreme actions, including “light electricity” and “moderate beatings,” are “permitted with negotiation,” though “body-shaming and race-shaming … are disallowed entirely.” The orgies have been known to attract a disproportionate number of employees at Anthropic and Ope…

daringfireball.net其他英文

How Microcredentials Can Help Engineers Build Trustworthy AI

Engineering teams are moving at breakneck speed to integrate artificial intelligence, but they are running into a formidable new barrier: compliance issues. Translating high-level AI ethics , governance , and auditing into daily operational workflows has become one of the industry’s toughest challenges. “AI ethics” refers to the principles developers use to govern and operate workflows. “Governance” refers to the rules and procedures an organization establishes for AI workflows and how they interact with organizational data. Auditing is the process of assessing and validating systems to ensure safety, fairness, transparency, and legal compliance. Technology leads and systems engineers are re…

IEEE Spectrum其他英文

Doctoral Advising Guidelines by the AHM

[This is a guest post by the AHM communications group. — T.] We are happy to announce that the Teaching Working Group of the Association for Human Mathematics wrote a set of recommendations on doctoral advising aimed at departments, advisors, and students. The document contains five guidelines centered on prioritizing human-authored writing, avoiding overreliance on […]

What's new其他英文

Can Nifty, Sensex extend gains after rallying up to 2% last week? 4 factors to watch on Monday

Indian benchmark indices Sensex and Nifty ended the week higher, gaining up to 0.5%, even as broader markets declined amid rising crude oil prices above $117 a barrel. Investors will track oil price movements, foreign institutional investor (FII) activity, global cues and domestic market trends for direction. FIIs turned net buyers after a 10-week selling streak, lending support to sentiment.

Economic Times财经资讯英文