Tweets

RT @_rockt: Our #ICLR2017 paper w/ @mdaniluk91 on key-value memory for neural language modeling is on arXiv! https://t.co/v5YIBNeipd #DeepL… — 3 days 10 hours ago
@omerlevy_ @nasrinmmm @yoavgo @haldaume3 we found the problem, i.e. bias. We should definitely address this rather than penalizing arXiving. — 3 days 10 hours ago
@yoavgo @haldaume3 @geomblog I don't think stop arXving is a solution though, and it is too late to stop it. Agree on the competitive argumt — 3 days 20 hours ago
@haldaume3 @yoavgo @geomblog I haven't seen short papers in TACL. Correct me if I am wrong. — 3 days 20 hours ago
@yoavgo @haldaume3 @geomblog I would not like to beef it up. Rather submit as a short paper to conference and move on. — 3 days 20 hours ago
@yoavgo @haldaume3 @geomblog Yes TACL was the best thing happened. Not sure if TACL likes incremental work. — 3 days 20 hours ago
@yoavgo @haldaume3 @geomblog this year NLP has 3 major deadlines all in 4-5 months. Not great for PhD students who missed these deadlines — 3 days 20 hours ago
RT @zehavoc: @haldaume3 @yoavgo @geomblog @sivareddyg @kchonyc @rctatman the way the imdb guys solved it was neat. https://t.co/nC7cwD2aTk… — 3 days 20 hours ago
RT @jayantkrish: @haldaume3 @yoavgo @geomblog @sivareddyg totally agree with this. ArXiv solves the "X depends on Y, but Y is unpublished"… — 3 days 20 hours ago
RT @haldaume3: @zehavoc @yoavgo @geomblog @sivareddyg @kchonyc @rctatman should we try to organize a debate at ACL? cd try to get ppl on al… — 3 days 20 hours ago
RT @geomblog: @yoavgo @sivareddyg @haldaume3 I think people maybe underestimate how much the arxiv had ALREADY changed publication culture — 3 days 20 hours ago
@yoavgo @haldaume3 @geomblog because the paper is out there, you can focus on other things or improve rather than pray. — 3 days 20 hours ago
RT @zehavoc: @haldaume3 @yoavgo @geomblog @sivareddyg @kchonyc @rctatman but yes, this is crucial question. How come there's no more pane… — 3 days 20 hours ago
@yoavgo @haldaume3 @geomblog no, conference acceptance should still be there but arXiv helps in prioritizing research, pursue multiple probs — 3 days 20 hours ago
@geomblog agree. Either 1) fully closed or 2) new measures for openness will work. But 1) is no longer possible and not a good model I think — 3 days 21 hours ago
@zehavoc @kchonyc @rctatman @haldaume3 Reddit & Twitter are addressing this partly, #SemanticScholar aggregates all media. More work TBD. — 3 days 21 hours ago
@zehavoc @kchonyc @rctatman @haldaume3 which is why wider audience should also be allowed to judge the merit of the work openly, like ICLR — 3 days 21 hours ago
RT @nasrinmmm: Congrats to Sap et al. @uwnlp for winning #StoryClozeTest shared task #LSDSem #EACL2017 Stay tuned for the report paper on o… — 3 days 21 hours ago
@zehavoc @kchonyc @rctatman @haldaume3 on the death of double-blind reviewing, I think this will happen more in fut… https://t.co/JPauiLoE9t3 days 21 hours ago
RT @brendan642: Mark Steedman throwing some comp ling shade (2011) https://t.co/TnlPsmpYWE #nlproc https://t.co/DE9Q6v6HMg4 days 7 hours ago
Site Counter