Kevin Jacobs

@[email protected] | Ph.D. candidate @MaastrichtU interested in how machines and people think, interpret and understand | #NLProc #IR

[Kevin Jacobs @kmjjacobs]

This post highlights research on automatic text summarization that I found interesting: https://t.co/9CJixO8sss #NLP #Summarization

[Anna Rogers @annargrs]

New blog post: BERT Busters👻: outlier dimensions that disrupt Transformers
https://t.co/HQLkuFKrVy
TLDR: We found the Achilles' heel of BERT&co! Disabling just <0.0001% of outlier layer output weights drops model quality this much:
/1

[Kevin Jacobs @kmjjacobs]

https://t.co/nW0MwhAlaP

[Shiyue Zhang @byryuer]

Hi #ACL2021nlp, check out our new work “EmailSum: Abstractive Email Thread Summarization” (11-12p ET Aug4 Wed 15D). 😀

w @real_Asli @JianfengGao0217 @MohitBan47 (@uncnlp+@MSFTResearch)

https://t.co/estXVP6b7N
Github: https://t.co/tBQJ2Nyru7
Video: https://t.co/enSOSBYpXV

🧵⬇️

[Jesse Vig @jesse_vig]

Is your summarization model hallucinating 👁️? Is it truly abstracting🧠? Or just extracting✂️?

Next week at ACL, stop by our demo of SummVis, an interactive vis tool to help you answer these questions.

Paper/code/demo: https://t.co/rGcb2KiIQu

@SFResearch @StanfordAILab

1/N

[Kevin Jacobs @kmjjacobs]

Finetuning BERT on Google Colab https://t.co/Rn5xagR6ga

[Philip Vollet @philipvollet]

Reconstructing Implicit Knowledge with Language Models.

Generating statements that explicate implicit knowledge connecting sentences in text.

https://t.co/QKz09sV62Z

https://t.co/wyGiWVcVkC

[Kevin Jacobs @kmjjacobs]

Simplifying PyTorch code using PyTorch Lightning https://t.co/hrAk5raBdj

[Alexandra Chronopoulou @alexandraxron]

Happy to share that our paper "Improving the Lexical Ability of Pretrained Language Models for Unsupervised Neural Machine Translation" got into #NAACL2021!🎉Joint work w/ @StDario1 and Alex Fraser.

Here is a summary thread👇

[Kevin Jacobs @kmjjacobs]

The start of a new year, the start of my new Research Blog https://t.co/jBCZVf5Gw9 - Happy 2021!

[Kevin Jacobs @kmjjacobs]

An interesting article on Pedestrian Detection using PyTorch by @sshkhr16 https://t.co/o7UmintjE4 #cnn #pytorch #pedestrian

[Kevin Jacobs @kmjjacobs]

How to Generate Tweets using GPT-2 https://t.co/QqCqD7nC1p

[Andrej Karpathy @karpathy]

Nice/fun YouTube channel walking through recent papers in deep learning in a video format, this episode on GPT-3. Cool! :)

[Kevin Jacobs @kmjjacobs]

Really enjoyed SIGIR '19, especially the Summarization & Information Extraction track. Thanks to all the orgaziners! #SIGIR #IR

[KDnuggets @kdnuggets]

Largest #primenumber ever found: 2^(77,232,917) – 1 is prime. has over 23 million digits https://t.co/Y1M1mwAw4A

[Kevin Jacobs @kmjjacobs]

@Fisher85M @JacBurns_Comext What about a quantum attack? https://t.co/QsNzdoOfZj #quantum #blockchain

[Raymond Hettinger @raymondh]

#python news: 😀 @gvanrossum just pronounced that dicts are now guaranteed to retain insertion order. This is the end of a long journey.

[Kevin Jacobs @kmjjacobs]

A great article (except for the screen-filling advertisements) on how AI is used successfully by Google and NASA for discovering 2 new planets. https://t.co/nO5z7m2PN8

[Kevin Jacobs @kmjjacobs]

A Character-based Seq2Seq model using Python and Tensorflow - Data Blogger https://t.co/HwRTSyIYtB #seq2seq #python #tensorflow #nlp #nltk #rnn #lstm

[Lee Baker @eelrekab]

"Models are opinions embedded in mathematics." Cathy O'Neill