Twitter | Search | |
Search Refresh
Sameer Singh Jul 11
In paper on SEARs w/ & , we find transformation rules that preserve meaning of sentences, but change predictions of systems! SEARs find bugs in TextQA, VQA, & sentiment analysis. Paper: , code: .
Reply Retweet Like
Ivan Vulić Jul 12
Our new survey on typological linguistics for NLP is available online. All that you ever wanted to know about modeling language variation and universals. Caution: it's a long read. :)
Reply Retweet Like
Ted Pedersen Jun 9
Reply Retweet Like
ACL2018 Jul 17
ACL now have a professional conduct committee to enforce anti harassment policy
Reply Retweet Like
Delip Rao Jul 9
New blog post: "Design Patterns for Production NLP Systems"
Reply Retweet Like
Delip Rao Jun 14
: Small word embedding nuance that's not often discussed: When using pre-trained embeddings, in many situations it is better to use the same tokenizer that was used to derive the embeddings.
Reply Retweet Like
sopan khosla Jul 16
NLP community needs this cheat sheet! The Hitchhiker’s Guide to Testing Statistical Significance in Natural Language Processing Day 2
Reply Retweet Like
Mohammed Jul 9
Two and half lines of review without any single (technical) word which could help to improve the paper..hard to believe I am reading reviews of a Rank A conference...Rank A conferences deserve Rank A reviewers...a sad state of affairs
Reply Retweet Like
Isabelle Augenstein Jul 17
business meeting: EACL will be held yearly
Reply Retweet Like
ACL2018 Jul 17
Replying to @acl2018
AACL will be every even year and the first one will be in 2020
Reply Retweet Like
EVALITA 2018 11m
We are glad to announce that will be giving a keynote at Stay tuned for details!
Reply Retweet Like
Emily M. Bender Jul 15
Thinking ahead to 2019, I think it might be good to have a tutorial on integrating ethics into the curriculum. Any folks interested in working on this?
Reply Retweet Like
Emily M. Bender Jul 13
In neural , when the LSTM or whatever is described as “reading” a sentence, does that mean word by word, character by character, or is it completely ambiguous if not specified as one or the other? (Asking about writing conventions here.)
Reply Retweet Like
Delip Rao Jun 15
NCRF++ : A neural CRF++ toolkit for sequence labeling tasks. Works pretty much similar to Taku's CRF++ package, but built with !
Reply Retweet Like

Related searches

@acl2018
Dirk Hovy Jul 9
Looking forward to ! If you are interested in postdoc positions in computational social science/ starting this fall, come talk to me.
Reply Retweet Like
ACL2018 4h
Best demo paper: Out of the box universal romanization tool
Reply Retweet Like
ACL2018 Jul 14
Deep reinforcement learning for NLP
Reply Retweet Like
ACL2018 Jul 16
Closing comments by Carolyn Rose: don't be afraid of difficult questions
Reply Retweet Like
Alex Conneau May 6
Our paper "What you can cram into a single vector: Probing sentence embeddings for linguistic properties" was accepted to ! We interrogate sentence embeddings about linguistic information. paper: tasks:
Reply Retweet Like
Pranav 7h
Alright, let's do some science.
Reply Retweet Like