Tweetovi
- Tweetovi, trenutna stranica.
- Tweetovi i odgovori
- Medijski sadržaj
Blokirali ste korisnika/cu @ukhndlwl
Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @ukhndlwl
-
Prikvačeni tweet
Excited to share new work!!! “Generalization through Memorization: Nearest Neighbor Language Models” We introduce kNN-LMs, which extend LMs with nearest neighbor search in embedding space, achieving a new state-of-the-art perplexity on Wikitext-103, without additional training!pic.twitter.com/hehcLnDaKz
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Urvashi Khandelwal proslijedio/la je Tweet
Enjoyed the kNN-LM paper by Khandelwal and Levy et al. (2019). Using an interpolated non-parametric and parametric model, they set a SOTA on Wikitext, reducing perplexity by 2.9 points. This approach helps with predicting long-tail language predictions. https://arxiv.org/pdf/1911.00172.pdf …pic.twitter.com/XUN4XarOlj
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Urvashi Khandelwal proslijedio/la je Tweet
.
@stanfordnlp people’s#ICLR2020 papers #1—@ukhndlwl and colleagues (incl. at@facebookai) show the power of neural nets learning a context similarity function for kNN in LM prediction—almost 3 PPL gain on WikiText-103—maybe most useful for domain transfer https://openreview.net/forum?id=HklBjCEKvH …pic.twitter.com/5yKRhhjZMr
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Our work on nearest neighbor language models has been accepted to
#ICLR2020 Woohoo!! Code coming in the new year!https://twitter.com/ukhndlwl/status/1191188235629711360 …
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Urvashi Khandelwal proslijedio/la je Tweet
Update: I could reproduce all NER results from the XLM-RoBERTa paper
https://twitter.com/_stefan_munich/status/1205465165828886528 …
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Urvashi Khandelwal proslijedio/la je Tweet
Monday 11/18: @ukhndlwl &@johnhewtt of@Stanford will share 2#NLP lectures: “Generalization through Memorization: Nearest Neighbor Language Models” & “Probing Neural NLP: Ideas and Problems”
Join us!https://bit.ly/2Xiggj1 Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Gains on cross-lingual benchmarks are amazing!!! And the 100 languages are clearly identified in the paper
#BenderRule Nice work@kakemeister and co!!https://twitter.com/kakemeister/status/1192491008006639616 …Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Urvashi Khandelwal proslijedio/la je Tweet
Super awesome work! Really nice results, especially on Domain Adaptation!https://twitter.com/ukhndlwl/status/1191188235629711360 …
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Urvashi Khandelwal proslijedio/la je Tweet
Generalization through Memorization: Nearest Neighbor Language Models Reduces ppl from 18.27 to 15.79 (sota) in Wikitext-103 using kNN and pretrained Wikitext LM without further training. https://arxiv.org/abs/1911.00172
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Urvashi Khandelwal proslijedio/la je Tweet
Improve your language model by converting it into a deep nearest neighbour classifier! The amazing
@ukhndlwl pushes SOTA on Wikitext-103 by nearly 3 points, without any additional training (and gets a few other surprising results too). https://arxiv.org/abs/1911.00172 https://twitter.com/ukhndlwl/status/1191188235629711360 …
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Work done at
@facebookai with amazing collaborators@omerlevy_,@LukeZettlemoyer and@ml_perception as well as my@stanfordnlp advisor@jurafsky!! Paper: https://arxiv.org/abs/1911.00172 Code available soon!Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
We also show that kNN-LM can efficiently scale up LMs to larger training sets and allows for effective domain adaptation, by simply varying the nearest neighbor datastore without further training. It seems to be helpful in predicting long tail patterns, such as factual knowledge!
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Urvashi Khandelwal proslijedio/la je Tweet
we talk about "interpretation methods" for neural models, and want our interpretations to be "faithful", but what does it really mean?
@alon_jacovi attempts to clear the mess and, points to where we are and where we should be going.https://twitter.com/alon_jacovi/status/1190019266915053571 …Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Urvashi Khandelwal proslijedio/la je Tweet
Excited to share our work on BART, a method for pre-training seq2seq models by de-noising text. BART outperforms previous work on a bunch of generation tasks (summarization/dialogue/QA), while getting similar performance to RoBERTa on SQuAD/GLUE
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Urvashi Khandelwal proslijedio/la je Tweet
A story about neural networks and language understanding with quotes from yours truly,
@sleepinyourhat,@omerlevy_,@annargrs, and others, plus an amazing illustration of BERT teaching neural networks to other BERTshttps://www.quantamagazine.org/machines-beat-humans-on-a-reading-test-but-do-they-understand-20191017 …Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Urvashi Khandelwal proslijedio/la je Tweet
"What's the Aquaman actor's next movie?" Complex questions are common in daily comms, but current open-domain QA systems struggle with finding all supporting facts needed. We present a system in
#emnlp2019 paper that answers them efficiently & explainably:http://qipeng.me/blog/answering-complex-open-domain-questions-at-scale.html …Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Urvashi Khandelwal proslijedio/la je Tweet
Depends on the number of “positive” examples in the dataset? Ok, I’m done.https://twitter.com/shivon/status/1183906372217257984 …
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Urvashi Khandelwal proslijedio/la je Tweet
Happening today! I am speaking about StanfordNLP, our new
#NLProc toolkit at@PyTorch Dev Conference. Come talk to me if you are around and interested!https://twitter.com/stanfordnlp/status/1182056782249648130 …Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Urvashi Khandelwal proslijedio/la je Tweet
in "Show Your Work," we look at the status quo in experimental reporting in NLP -- it's abysmal -- and propose concrete ways to do better. https://arxiv.org/abs/1909.03004 to appear at EMNLP, by
@JesseDodge,@ssgrn,@dallascard,@royschwartz02, &@nlpnoahHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Urvashi Khandelwal proslijedio/la je Tweet
New
#emnlp2019 paper alert: So, there are quite a few methods for trying to uncover what an NN model _knows_ about some task. If you ask the same question several different ways, will you get the same qualitative conclusion? (1/N) https://arxiv.org/abs/1909.02597Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
Čini se da učitavanje traje već neko vrijeme.
Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.
coming soon...