Il faut mettre de toute urgence des cours de programmation dès le collège. Pas besoin de former des prof qui seraient de toute façon jamais au niveau. Il suffit de s’organiser pour apprendre aux enfants à coder en utilisant ChatGPT. Et ChatGPT c’est le prof. 1h/semaine.
LLMS
-
Join VISION IA Community for AI Learning and More
By
–
Merci d'avoir lu ! Je vous apprends ça et bien plus. Pour apprendre à utiliser l'IA, rejoignez les plus de 2000 participants qui ont déjà commencé ! Apprenez l'IA sous toutes ses formes et rejoignez la communauté VISION IA !
-
Understanding Transformers via N-Gram Statistics: Physics for Language Models
By
–
the closest thing i've seen to actual "physics for LMs" was probably this (single-author!) paper from neurips 2024: Understanding Transformers via N-Gram Statistics this is how we used to think about LMs; not sure why we stopped.
-
Questioning the Fundamental Definition of AI Agents
By
–
Yeah I don’t think “agents” is used in this way but … it feels a bit wrong. What’s some actual fundamental distinguishing property wrt what already exists? “It happens to use an LLM somewhere” is I think the current usage but imo it’s kind of lame.
-
Are LLMs Plateauing? Scaling Laws Running Out of Gas
By
–
Are LLMs plateauing? Are Scaling Laws Running Out of Gas? Many think so because: Data ceiling – We’ve scraped most of the internet. Synthetic data helps…but models aren’t great at grading their own homework. Compute pain – Transistors are approaching atomic-scale; energy
-
Text-to-SQL Converter with LangChain and Ollama Tutorial
By
–
Text-to-SQL Tutorial Build a powerful natural language to SQL converter using LangChain, Ollama's DeepSeek model, and Streamlit. This tutorial shows you how to create an intuitive interface that automatically converts spoken queries into database-ready SQL. Watch now:
-
Open Source Collaboration Accelerates AI Progress and Innovation
By
–
Without Llama (or PyTorch) there would have been no DeepSeek.
The point of open source is that you can accelerate progress by building on each other's work. -
Hidden Economic Knowledge in LLM Embeddings Revealed
By
–
Revealing economic facts: LLMs know more than they say This paper shows that hidden states (embeddings) of large language models (LLMs) contain rich economic information that can be used to estimate and impute economic and financial statistics more accurately than the LLMs’ text
-
Maya: Open-Source Multilingual Vision Language Model
By
–
Behind Maya: Building a Multilingual Vision Language Model This paper introduces Maya, an open-source multilingual Vision-Language Model (VLM) designed to enhance performance on vision-language tasks across eight diverse languages. Maya addresses the underperformance of existing
-
MiMo-7B: Reasoning-Optimized Language Model Outperforms Larger Models
By
–
MiMo: Unlocking the Reasoning Potential of Language Model — From Pretraining to Posttraining MiMo-7B is a reasoning-centric LLM series optimized from pretraining through posttraining, achieving state-of-the-art performance on math and code tasks—even outperforming models 4× its