开启时会模糊预览图，关闭后正常显示
ChatGPT 爱用破折号和擅长编程，这不是文科生和理科生的差别，更像是文科学习和理科学习的差别，当然实际上是两种不同类型内容生成的训练方式差别。 生成文本的结果是没有标准答案的，这就依赖于人类定义好的标准答案和反馈，偏好极其依赖于标注员标注的结果，就好比我们学文科，会依赖于我们阅读的教材和老师的讲解。 写代码和做数学题的特点是有标准答案，就好像我们学理科做数学物理，对就是对，错就是错，和人类偏好无关，做对了就有奖励，做错了就没有奖励，所以模型能反复训练，最终越来越厉害。
Prompt Engineer, dedicated to learning and disseminating knowledge about AI, software engineering, and engineering management.
👀
modeling language at @allen_ai
RT @okooo5km: 去年 WWDC 我最期待 Apple Intelligence，但到现在仍没有让我眼前一亮的系统体验；也正因此，我把目光落在更务实的接口上——AppIntents。名字起得很好：Intents 是意图，用户的需求就是意图；AI 必须听得懂这种意图，也…
🔖 Building · Sharing ✨ Zipic · TimeGo · KeygenGo · HiPixel · Orchard · Prism
自从决定创业以后， 工作对我来说就是一种奖励。
Growth Coach｜Helping creators build their personal brand on X wechat official：PandaTalk8
I just thought about how the OpenAI people on Twitter made those posts around the time of Altman’s ouster that included heart emojis of different colors. I remember thinking at the time that the colors must signify something because it seemed weirdly coordinated. It just occurred to me that perhaps they were indicating by the color which team they were on: Altman, Mira, or Ilya. If so, it would be a quite clever game theoretic approach, because it would be public and binding and would indicate their preferences to each other, all while looking like generic sentiments of support to the outside world. Someone should go back and try to cross reference those emoji colors to see which of those people subsequently stayed at OpenAI, which ones went with Ilya to SSI, and which ones went with Mira to ThinkingMachines.
Former Quant Investor, now building @lumera (formerly called Pastel Network) | My Open Source Projects: https://t.co/9qbOCDlaqM
RT @mmitchell_ai: Anyway, hi @X, I'm back. Was in Bsky but its moderation is (currently) so poorly managed that a mass amt of people I resp…
Co-founder & CEO @HuggingFace 🤗, the open and collaborative platform for AI builders