New research: FlashAttention-4 FlashAttention-4 achieves up to 1.3x speedup over cuDNN 9.13 and 2.7x over Triton on B200 GPUs with BF16. FlashAttention-4 co-designs algorithms and kernel pipelines for Blackwell GPUs, where tensor core throughput doubles but memory bandwidth and
RESEARCH
-
Marimo Thermodynamic Linear Algebra Framework for ML
By
–
marimo | thermodynamic_linear_algebra – marimo-team/gallery-examples https://
buff.ly/tXIcg0o
#AI #MachineLearning #DeepLearning #LLMs #DataScience -
A Look at Cortical Labs and Biological Computing
By
–
Here is another look at Cortical Labs with its founder: https://t.co/rLgoYMtrPh
— Robert Scoble (@Scobleizer) 7 mars 2026Here is another look at Cortical Labs with its founder:
-
Exploring biological computing with Cortical Labs
By
–
Yeah, so amazing tech.
— Robert Scoble (@Scobleizer) 7 mars 2026
I was first to show the world Cortical Labs and the interview with its founder is still great to understand what he's building: https://t.co/KaTj5AzyQtYeah, so amazing tech. I was first to show the world Cortical Labs and the interview with its founder is still great to understand what he's building:
-
Contest Tests AI-Ready Teaching Methods and Innovation
By
–
Contest puts AI-ready teaching to the test #AI #AIio #AIInnovation #ML #DataScience #Futureofwork @Scobleizer @AndrewYNg @drfeifei @KirkDBorne @fchollet @rowancheung @antgrasso ow.ly/jJiw30sTyF2
→ View original post on X — @terence_mills, 2026-03-07 19:00 UTC
-
QVGen enables compressed video generation models for standard hardware
By
–
Can we finally run massive video generation models on standard hardware without sacrificing quality? HKUST, Beihang, SenseTime Researchand others present QVGen. They created a training method that uses temporary helper modules to stabilize the model as it is compressed down to
-
Large Concept Models: AI Beyond Token-Level Language Models
By
–
Large Concept Models (LCM): A New Frontier in AI Beyond Token-Level Language Models https://
linkedin.com/pulse/large-co
ncept-models-lcm-new-frontier-ai-beyond-giuliano-liguori–dnj3f
… via @ingliguori -
AI Detection Tools: Do They Really Identify Deepfakes Effectively?
By
–
These Tools Say They Can Spot A.I. Fakes. Do They Really Work? https://
buff.ly/STIR3Wy
#AI #MachineLearning #DeepLearning #LLMs #DataScience -
Retrieval, Not Writing, Boosts AI Memory Accuracy
By
–
most people building AI agents obsess over how they WRITE memories.
turns out that's basically irrelevant. new research analyzed 9 different memory systems across 1,540 questions.
the finding? – retrieval method drives 20-point accuracy swings.
– write strategy? 3–8 points max. -
Research Paper on Consciousness in Artificial Intelligence
By
–
Why Is Anything Conscious? Michael Timothy Bennett, Sean Welsh, Anna Ciaunica: https://
arxiv.org/abs/2409.14545 #ArtificialIntelligence #AIAgents