所以我做了一个 50 秒的演示视频,带大家把海外刷屏的 JEV 以及它背后的底层架构看透👇(见下方视频)
This is a AI post classified by Jev as Frontier models (a model release), kept by the AI Radar because it carries real work, not commentary.
很多人没看懂爆火的JEV是什么? 所以我做了一个 50 秒的演示视频,带大家把海外刷屏的 JEV 以及它背后的底层架构看透👇(见下方视频) JEV 是前 OpenAI、参与 ChatGPT / RLHF 的 @CompleteSkeptic(Diogo Almeida)创办 @typesafeai 后的第一款模型 他在 OpenAI 时发现了一个致命矛盾: 通过人类偏好微调(RLHF)训练出来的自回归大模型,存在严重的模式崩溃(Mode-dropping)、过度自信与幻觉,且串行逐字生成的延迟极大,根本无法胜任严肃的机器自动化生产 为此,他们花了两年研发了专为自动化设计的 System 1 基础模型 JEV,核心架构创新非常激进: • 单次并行计算(One Parallel Pass): 如同当年 Transformer 取代 RNN 一样,JEV 彻底抛弃了逐 Token 串行吐字的自回归机制,采用全新的模型架构与采样器,一次性并行击发所有预设选项 • 强化学习校准决策(RL for Calibrated Decisions): 引入全新的训练算法,输出媒介不再是自然语言文本,而是直接输出经过置信度校准的概率分布 • 代码级可靠(TypeSafe): 彻底根除幻觉,输出自一致、确定性高,可作为强类型对象直接被程序调用 • 100 倍提速与极端降本: 延迟降至 0.
Posted by 梭哈.AI (32.7k followers) 1 days ago · 223 likes · 425.4k views · view the original post on X. Kept by the AI Radar as Frontier models.
More AI work like this
- TikTok reacts to this: — @_AashishReddy
- Lyria 3.5 from Google DeepMind is now live on fal — @fal
- Tested Jev vs GPT-5.6 for predicting my taste in books. — @venturetwins
- Wife wanted 5-yo involved, so Fable gave me a bunch of reading questions to give him to… — @kevinnbass
- opus 5.2 learned to write poems — @aliceisplaying
- meanwhile grok first try: — @thekitze
- Les premières pièces totalement modélisés par GPT6-Astra — @DFintelligence
- gpt 5.6 terra = gpt 6 astra — @notjazii
Every post is read and classified by Jev (TypeSafe): what it is, which market it belongs to, and whether the link is a real tool. 31.3k posts from 4.7k X accounts over the last 14 days, 1.3k tools, 19 markets. Collected every 5 minutes, fully re-ranked every hour — last update 2026-09-19 21:10 UTC. Full method.