Ran Levinstein

Accelerate your transformer model with the new Block-Sparse-Flash-Attention! github.com/Danielohayon/B… This training-free, drop-in replacement extends FlashAttention-2 with minimal code changes (CUDA Kernels Included). Paper: arxiv.org/abs/2512.07011

[1/6] Tomorrow (Thursday) at #NeurIPS: Are Greedy Task Orderings Better Than Random in Continual Linear Regression? Q: Do models learn better when consecutive tasks are similar or dissimilar? A: Our analysis suggests that they should be dissimilar! openreview.net/forum?id=8JdPq…

[1/5] Next week at #NeurIPS *Optimal Rates in Continual Linear Regression via Increasing Regularization* In the brain, ageing naturally reduces synaptic plasticity. Our theory suggests continual learning models may benefit from a similar mechanism! openreview.net/forum?id=lDh78…