Revisiting Character-Based Neural Machine Translation with Capacity and Compression, from @GoogleAI. “We show that deep models operating at the character level outperform identical models operating over word fragments.” https://t.co/RR7GOI2ku4
— hardmaru (@hardmaru) August 30, 2018
Prediction: By the end of this calendar year, it will be so easy to run a boosted or neural network machine-learning model using Twitter data (like the one below) that we social science folk are actually gonna have to start learning it [so we can review/respond to reviewers]. https://t.co/zdlYXFbrdY
— Mike Kearney📊 (@kearneymw) August 29, 2018
A very nice paper for those interested in Transformers for NLP (and if you are not, you should!) Give insights on why these models improve on high level metrics like BLEU/ppl. I like that they went the extra mile to get a real comparison to the (interesting) results of @ketran! https://t.co/9qoOWArtlD
— Thomas Wolf (@Thom_Wolf) August 29, 2018
They trained a model to generate the model parameters of the encoder and decoder LSTMs used for multilingual machine translation! https://t.co/B5kTkZ00HD
— hardmaru (@hardmaru) August 28, 2018
#EMNLP2018 paper by @eaplatanios on Context Sensitive Parameter Generation for Universal NMT! https://t.co/93mLfECxnC
— Graham Neubig (@gneubig) August 28, 2018
Inspired by @hardmaru's HyperNetworks, we learn to generate NMT parameters for the languages we want to translate. Nice results on multilingual and zero-shot MT! pic.twitter.com/QOYrSwoPMq
A new #deeplearning model for improving abstractive summarization by actually creating novel phrases. Reinforcement learning in #naturallanguageprocessing. Work by amazing intern @iam_wkr and Salesforce Researchers Romain Paulus and @CaimingXiong.https://t.co/gXcp3gyOov pic.twitter.com/UBK8V0VDC6
— Richard (@RichardSocher) August 28, 2018
Check out our intelligent @Slack bot named Mr. Jingles! Our office mascot, once a humble cat, can now fetch stories, create visualizations, and even explain spikes in news coverage :smiley_cat: #mediamonitoring #chatbots https://t.co/HcZRIX13G7 pic.twitter.com/lsZ98oRU23
— AYLIEN (@_aylien) August 27, 2018
"Style Transfer as Unsupervised Machine Translation," Zhang et al.: https://t.co/aztJfKyBpc
— Miles Brundage (@Miles_Brundage) August 27, 2018
No one cites anything from before 2013 these days, of course—except Hochreiter and Schmidhuber—but really the overall approach here recalls Answer Validation, an original leading motivation for RTE/NLI. https://t.co/dMnCirMJWY – https://t.co/SrinQwyP6n https://t.co/3H63d3cnpy
— Stanford NLP Group (@stanfordnlp) August 26, 2018
NLP.js – node.js NLP library for building chat bots, with entity extraction, sentiment analysis, automatic language identify, and so more.https://t.co/o1uxvKkujW #MachineLearning pic.twitter.com/ekzznbvzLl
— ML Review (@ml_review) August 25, 2018
Interesting on why state of the art NLP still has a long way to go https://t.co/Ml1TxbTJnC
— Erik Bernhardsson (@fulhack) August 23, 2018
People usually get information from others in a multi-turn conversation. To approach this, we’ve released CoQA 🍃—A Conversational Question Answering Challenge by @sivareddyg•@danqi_chen•@chrmanning. 127K Qs— free-form answers—with evidence—multi-domain. https://t.co/XkdWC2Hh3W pic.twitter.com/v0FXMnNg8n
— Stanford NLP Group (@stanfordnlp) August 22, 2018
