I’m co-organizing an “AI for Science: Algorithms to Atoms” social event during #NeurIPS2025 with Yann LeCun, Anima Anandkumar, Bill Dally, and Max Welling If you want to talk about AI Scientist, World Models and the future of AI-driven discovery, please come on Dec 5 3:30pm PT!
MACHINE LEARNING
-
Model outputs text and images with visible CoT
By
–
There’s text CoT (visible in the AI Studio UI) but my understanding is it’s from a single model that outputs both text and images
-
Keras 3 JAX Backend: Easy Adoption with KerasHub Integration
By
–
If you want an easy way to adopt JAX, use Keras 3 with the JAX backend. Need pretrained models? KerasHub gives you access to any model from Hugging Face. All running with native JAX performance.
-
JAX: A Competitive Advantage You Can’t Ignore
By
–
If you're not using JAX, you're missing out on a huge competitive advantage.
-
Fundamental Limits of LLMs at Scale: Mathematical Foundations
By
–
10. Fundamental Limits of LLMs at Scale Establishes mathematical foundations for theoretical limitations constraining LLMs: hallucination, context compression, reasoning degradation, retrieval fragility, and multimodal misalignment.
-
LAMP: Language-Augmented Multi-Agent Reinforcement Learning Pipeline
By
–
9. LAMP: Language-Augmented Multi-Agent RL LAMP integrates natural language processing into multi-agent reinforcement learning through a three-stage pipeline.
-
How Agentic AI Works: From Data Collection to Autonomous Action
By
–
How does Agentic AI work? Collects data (APIs, sensors, logs, queries) Reasons & plans with memory + context Executes actions, learns from feedback Produces autonomous, adaptive responses AI that doesn’t just respond — it acts.
#AgenticAI #AI #Automation -
EGGROLL: Gradient-Free LLM Training Achieves 100x Throughput
By
–
Could evolution replace gradients for LLM training? Introducing EGGROLL, a new paper that makes gradient-free training actually practical 100x training throughput vs vanilla ES 100x memory reduction 91% inference speed
Pure integer training Trending #1 on alphaXiv -
Gemini 3 Secret: Pre-training and Post-training Improvements
By
–
Actually if you want to know what the real ‘secret’ is 😀 it’s world-class research AND world-class engineering AND world-class infra all working closely together with relentless focus and intensity… Oriol Vinyals (@OriolVinyalsML) The secret behind Gemini 3? Simple: Improving pre-training & post-training 🤯 Pre-training: Contra the popular belief that scaling is over—which we discussed in our NeurIPS '25 talk with @ilyasut and @quocleix—the team delivered a drastic jump. The delta between 2.5 and 3.0 is as big as we've ever seen. No walls in sight! Post-training: Still a total greenfield. There's lots of room for algorithmic progress and improvement, and 3.0 hasn't been an exception, thanks to our stellar team. Congratulations to the whole team 💙💙💙 — https://nitter.net/OriolVinyalsML/status/1990854455802343680#m
→ View original post on X — @oriolvinyalsml, 2025-11-22 20:32 UTC
-
Les sept sommets de l’intelligence artificielle : évolution et stratégie
By
–
Scaling Ascent Peak: The Seven Summits of Artificial Intelligence In my latest article, I chart the evolution of AI—from the early days of symbolic logic to today’s autonomous agents, and what lies beyond. This journey is more than a history lesson: it’s a strategic framework