By
–
Liquid AI released LFM2-2.6B-Exp, a small, open 3B model that outperforms DeepSeek R1-0528 in certain benchmarks, such as IFBench. I believe that after AGI, we will have a 3B open-source model with the same performance, which can run on any device. Small intelligence
By
–
BREAKING : OpenAI released GPT-5.2-Codex-XMas model which performs exactly as GPT-5.2-Codex. It should be crushing on Santa bench
By
–
6/ Simulate real exams or real work
Theory without pressure is useless. Prompt:
“Create a realistic exam or real-world task that tests whether I actually understand this topic. Grade my response harshly.” Painful. Effective.
By
–
2/ Force depth, not vibes
AI loves being surface-level unless you constrain it. Prompt:
“For today’s lesson, teach the concept as if you’re a university professor who is not allowed to simplify or skip steps.” The quality jump is wild.
By
–
SHOCKING: I stopped taking online courses. ChatGPT now builds me custom curriculums for anything I want to learn. Here are 7 prompts that turned it into a personal professor
By
–
le high et xhigh c'est juste incroyable. Bon par contre ca prend 40min mais c'est un truc de fou.
By
–
running automations without context about your business is like running a frontier AI model on a nokia phone you can fix that with context profiles i added context profiles inside our automations bundle → AI interviews you about your business → generates JSON profile (ICP,
By
–
Windsurf released Wave 13 update with upgraded parallel agents, dedicated terminal and extended Free trail for SWE-1.5 model.
By
–
Groq has entered into a non-exclusive licensing agreement with Nvidia for Groq’s inference technology. GroqCloud will continue to operate without interruption. Learn more here: groq.com/newsroom/groq-and-n…
By
–
‼️VLMs/MLLMs do NOT yet understand the physical world from videos‼️ In our recent work, we found that even the most advanced AI models still lag behind humans in one key aspect: reasoning about the kinematic properties of objects from videos. Takeaways: 1. ChatGPT 5.1 leads overall among 21 advanced VLMs, followed by Gemini 2.5 Pro/Flash. 2. Grok 4.1 delivers impressive performance at the lowest API cost. 3. Qwen3-VL is the top-performing open-source model. Read here: quantiphy.stanford.edu/ 🧵1/N