Explore
Newest first — browse tweet threads
Keep on to blur preview images; turn off to show them clearly
Delphi is our biggest release yet. It’s the market for machine intelligence, allowing you to express your view on the best models, and win if you’re right. We now have a live tracking price for the best models. 📈📉
COO @GensynAI / ex @a16z @Cravath
实测了智谱新上新的GLM4.6V，从放出的结果上看，GLM-4.6V 指标跟 Qwen3-VL-235B 持平，在z. ai上输出内容的时候可以引用长文档不同位置的图片截图，输出图文混合的排版👇
指标在这👇
如何获得大量bad case的案例： 先收集一下网络各种脏话和抱怨 训练一个识别agent锁定session片段 把session提取出来 这样就拿到了badcase… 也就是说 如果你能做一个用脏话识别badcase的agent 这东西可能也有市场 因为大部分人懒得做这么一个东西去识别bad case…..
Believing is seeing
"Please explain, in English, why this code is wrong." Opus 4.5: the problem is <wrong explanation> Gemini 3: <correct code> I don't get it, how can Gemini 3 be smart enough to find issues that Opus 4.5 failed to, yet not able to understand a simple instruction. LLMs...
wait this post has actual humans commenting on it hi everyone, missed you 🥺
Nano Banana 2 Flash 图像模型即将发布 图像质量相当不错，应该是 Gemini 3 Flash 驱动，看起来会便宜不少 Pro 确实太贵，谷歌都不敢免费了，可能 Nano Banana 2 Flash 谷歌就能免费了 而且其他产品的免费额度也能提高，太期待了
关注人工智能、LLM 、 AI 图像视频和设计（Interested in AI, LLM, Stable Diffusion, and design） AIGC 周刊主理人｜公众号：歸藏的AI工具箱
The people fighting me for saying not to all votes should count equally are the same ones asking why we are conducting airstrikes on a coup on our doorstep.
Founder | Author | Speaker Building @beltstripe. Healtech/EdTech/Agric I'm Not The Man Of Your Dreams. Your Imagination Wasn't This Great.