A few of us took a look at theoretical limits of large language models (LLMs) as they scale up to larger sizes. The bottom line is that larger size alone is not sufficient to address some fundamental concerns such as hallucination. #LLMs #CS #AI #ML

Enjoy!

openreview.net/pdf?id=BIRDG...

@markhamillofficial.bsky.social this seems forced.

I'm very proud of my PhD advisee Naeem Shahabi Sani and his co-authors Ferial Najiantabriz and Shayan Shafaei for earning the award for the best neuroevolution paper at ACM GECCO 2026! #GECCO2026 #GECCO #EvolutionaryComputation #MachineLearning #NeuroEvolution

Actual photo of me heading to CVPR last month. #CVPR2026

Could/should. What’s the difference?

Crimson Permanent Assurance spotted near OU campus.

Attending CVPR 2026 in Denver, it appears to me that AI research workshops are on the brink of extinction. Can and should we save them?

#CVPR2026 #AIResearch