开启时会模糊预览图，关闭后正常显示
Gemini 3 Flash 现已：成为 Gemini App 和 AI Mode in Search 的默认模型。 详细介绍：https://t.co/6ytSrhLuWz
艹 不科学啊 在MMMU-Pro多模态理解基准测试中 Gemini 3 Flash 竟然超越了Gemini 3 Pro😅
RT @TacoCohen: People keep running into this kind of bug in LLM-RL codebases, and it's exactly why I wrote our Env API from the get go to 1…
achieve ambition with intentionality, intensity, & integrity i made: - @dxtipshq - @cognition - @sveltesociety - @aidotengineer - @latentspacepod + @smol_ai
RT @iwsfutcmd: there's an old joke: a man meets an old woman, is making conversation, and asks her where she's lived in her long life. she…
Choose disfavour where obedience does not bring honour. I do math. And was once asked by R. Morris Sr. : "For whom?" @halvarflake@mastodon.social
what do openai and anthropic even do rn man? there is no way they keep charging 5x for a sub 5% gain at 5x the latency.
Prompt engineering reality🪄🧠 ✨Building AGI for work✨
兄弟们，炸裂了 Google 发布 Gemini 3 Flash 模型 Pro 级智能 × Flash 级速度 具备博士级逻辑与推理能力 性能超越Gemini 2.5 Pro，同时速度快 3 倍 成本仅为其零头 它能根据问题复杂度「自己调节思考时间」—— 简单问题秒答，复杂任务多想几步。 完成同等任务时平均使用 30% 更少 tokens 该模型适用于： 实时交互式应用（如语音助手、游戏、教育工具） 快速编码与分析任务 高并发 API 请求环境 低延迟、多模态 AI 应用
具备博士级逻辑与推理能力 在复杂问题、多模态理解、代码生成等任务上与大型模型相当。
Gemini 3 Flash across different test-time compute levels (green line below) represents a new score/cost Pareto frontier on ARC-AGI-2. Congrats to @demishassabis and @sundarpichai on the launch!
Co-founder @ndea. Co-founder @arcprize. Creator of Keras and ARC-AGI. Author of 'Deep Learning with Python'.
