François Chollet on X: "Meanwhile, AGI will in fact get better by simply adding more *compute*. It will not be bottlenecked by the availability of human-generated text." / X
MLPerf Introduces Largest and Smallest LLM Benchmarks
How o3 and Grok 4 Accidentally Vindicated Neurosymbolic AI
Think Only When You Need with Large Hybrid-Reasoning Models
View PDF
Report: Spring 2025 AI Model Usage Trends - Poe
Measuring AI Ability to Complete Long Tasks
(Halfway there, like Zeno. Invites proxies.)
Inception emerges from stealth with a new type of AI model | TechCrunch
How is Google using AI for internal code migrations?
View PDF
Titans: Learning to Memorize at Test Time
What just happened
Model-Based Transfer Learning for Contextual Reinforcement Learning
New Computing Breakthrough achieves 100 Million Times GPU Performance
Learning and performance in the world of generative AI
Japanese researchers set new world record internet speed
Google Sheets doubles calculation speed in Chrome, Edge
Scalable MatMul-free Language Modeling
View PDF
Nvidia Conquers Latest AI Tests
Training Compute of Frontier AI Models Grows by 4-5x per Year
Training Compute of Frontier AI Models Grows by 4-5x per Year
Evaluating Large Language Models Using “Counterfactual Tasks”
anton on X: "This cost + latency for llama 3 is actually insane. Just look at the rest of the models in comparison https://t.co/J1GKQSP5EA" / X
Intel says Copilot will run on your laptop — but only Snapdragon chips can handle it
Simultaneous and Heterogenous Multithreading | Proceedings of the 56th Annual IEEE/ACM International Symposium on Microarchitecture
PDF
The GPT-4 barrier has finally been broken
The Fastest AI Chip in the World Explained
Meet 'Groq,' the AI Chip That Leaves Elon Musk’s Grok in the Dust
Groq
‘Scalene’ offers efficiency suggestions for Python coders
A new quantum algorithm for classical mechanics with an exponential speedup
New AI Model Outperforms Google's Powerful PaLM-2
New techniques efficiently accelerate sparse tensors for massive AI models