Joseph Viviano

whatsinthebox. analysis of public NLP corpora

gflownet-tutorials. Interactive demos describing the main concepts behind GFlowNets

wae. ift6135 final project

epitome. scriptit modules for fmri analysis

dis-ult-diff-mod. dishonored ultimate difficulty mod

lstm-variational-dropout. theoretically grounded application of dropout to LSTMs

pyddsm. python data extraction tool for the ddsm project (digital database for screening mammography)

ift6135. learning representations

ift6390. machine learning

xbrain. classify individuals into cognitive groups, or biotypes (brain/behaviour)

saliency-red-herring. Code for Saliency is a Possible Red Herring When Diagnosing Poor Generalization, Viviano et al ICLR 2021

intro-dl. Some familiarity with MLPs

biotype. Python

qcmon. qćmon: gotta check 'em all

mdmr. functions for doing multivariate regression analysis of distance matrices

comp550. natural language processing

interview-basics. little bits of code i write so i don't forget how to walk

vae-latent-constraints. train a vae unsupervised, then train an actor critic to apply constraints to the latent space using parse tags from the penn treebank

trn. code related to viviano & schneider 2015, journal of neuroscience

resume. my credentials

pycaso. Python

transfer-learning-google-quick-draw. madamne ricard and the booster boiz

humanware. Jupyter Notebook

intraday_nope_research. Julia

adaboost-capacity-eval. madamne ricard and the booster palz investigate whether the main benefit to adaboost is adding capacity to the ensemble, vs. the introduction of models that specialize on subsets of the data

omsignal. information: https://github.com/mila-udem/ift6759/tree/master/projects/omsignal/slides

eqprop. Equilibrium Propagation

groundedSCAN. Grounded SCAN data set.

brainhack-dl. brainhack 2019 deep learning teaching material

horoma. https://github.com/mila-iqia/ift6759/tree/master/projects/horoma/data

sysadmin. linux sysadmin course