LLMs still struggle with long theorem proofs. LeanProgress offers a solution. We introduce a critic model for Lean4 proof search, and it uses a distance-based measure. Instead of just log-probability (logp), our critic uses “distance” as a signal, guiding the LLM to explore and
LLMS
-
GlobalGPT lets you chat with multiple large models
By
–
🚀 Ce que vous pouvez faire avec GlobalGPT :
— Jouhatsu | AI Influence Operator (@Jouhatsu_ai) 3 mars 2025
1- Chat IA
Discutez avec ChatGPT-4.5, DeepSeek-R1, Claude 3.7 Sonnet, Gemini 2.0, Llama 3.3 et d'autres modèles puissants.
Plongez au cœur de leurs capacités de raisonnement. pic.twitter.com/SwrBCG6OmRCe que vous pouvez faire avec GlobalGPT : 1- Chat IA Discutez avec ChatGPT-4.5, DeepSeek-R1, Claude 3.7 Sonnet, Gemini 2.0, Llama 3.3 et d'autres modèles puissants. Plongez au cœur de leurs capacités de raisonnement.
-
Mastering ChatGPT: The Art of Effective Prompt Engineering
By
–
We started with the fundamentals: crafting effective prompts for ChatGPT. Turns out, talking to AI is an art form! The more specific and structured your request, the better the results you'll get. It's like learning a new language that speaks back with incredible insights.
-
Focus on Writing About Reasoning Models Over Video Content
By
–
Good question. Currently, I am more focused on writing about reasoning models (vs making videos on reasoning models) because the research landscape if evolving so quickly. Probably I'll make one in the future (more likely 2026ish) once the methodology converged.
-
New Python Course Combines Coding with Generative AI
By
–
A tough decision we have to make daily… Why not do both? We have a new course teaching Python leveraging LLMs the right way If you’ve been looking for a jumpstart to combine coding with state-of-the-art AI, we’ve got you covered. The Python Primer for Generative AI is your
-
Companies Developing LLMs for Code Generation
By
–
Companies specifically focusing on developing LLMs for code generation. There are the obvious ones, but I also know a couple of startups who are laser-focused on this.
-
LLMs struggle with TikZ figures but code generation training improves automation
By
–
Yes, LLM also still struggle with Tikz-generated figures etc. I can imagine though that with all the focus on code-generation training, this will be easier to automate.
-
AI Model Competition Accelerates: GPT-4.5, Sonnet 3.7, Grok 3
By
–
In 2023 every week in AI felt like a year. The acceleration is even more rapid now with competition. In the last 2 weeks alone we got: + GPT-4.5, the best model for chat + Sonnet 3.7, the best model for code
+ Grok 3, the most fun model … and o3-mini was only a month ago -
Groq AI Weekly: GroqCloud Updates, Rate Limits, Mistral Saba
By
–
The latest edition of Groq AI Weekly covers GroqCloud self-serve, rate limit increases, our Mistral Saba model drop, Orq AI’s, end-to-end GenAI collaboration platform powered by Groq, and more. https://
hubs.la/Q038W92G0
Subscribe on YouTube to keep up with our news each week.
