Hi! I’m Vlad, and I’m a postdoc within DeepSPIN at the Instituto de Telecomunicações, Lisbon, Portugal. I research machine learning & natural language processing, focusing on sparsity and structure.
I obtained my PhD from Cornell in 2018, advised by Claire Cardie. I am a scikit-learn core dev/alum and I develop the polylearn library for factorization machines and polynomial networks in Python.
News
-
Released dynet code for our EMNLP‘18 paper, Towards dynamic computation graphs via sparse latent structure.
-
SparseMAP: Differentiable sparse structured inference is accepted at ICML‘18! PyTorch implementation (dynet coming soon).
-
Released code for our NIPS‘17 paper on A regularized framework for sparse and structured neural attention.
Contact
You can find me on twitter as @vnfrombucharest or on github as @vene. I’m also on facebook and g+. Or just e-mail me: vlad at vene.ro. My public key is available from Keybase.