@NYUDataScience
Pinned post
nyudatascience.medium.com A New Tool Gives Researchers Evidence to Back Their AI Training Claims A researcher can publish an AI model’s code and results and still leave readers unable to check how the model was actually trained. CDS… 
www.linkedin.com Paper Accepted to NeurIPS 2026 on Memorization in Image Classification | Jeevesh Attri posted on... Following up on a project I shared earlier, I’m thrilled to say that our paper, “Memorization in Image Classification: Test Set Behavior and Scaling Laws,” has been accepted to the NeurIPS 2026... 
nyudatascience.medium.com CDS Researchers Challenge Anthropic’s Evidence That Language Models Can Introspect In 2025, Anthropic reported that its Claude models could detect when researchers injected a concept directly into their neural activity… After over a century of studying introspection in humans and animals, researchers are beginning to explore the phenomenon in unlikely new subjects — large language models. These systems, which are based on mathematical structures called neural networks, excel at cognitive tasks that until recently seemed exclusive to humans. Can they introspect as well? @benbenbrubaker reports: quantamagazine.org/new-research-t…View quoted post on X
