Papers of the day   All papers

EXBERT: A Visual Analysis Tool to Explore Learned Representations in Transformers Models

Comments

Ben Hoover: Pleased to announce exBERT, an interactive tool to explore the embeddings and attention of Transformer models at different layers, with different heads. Demo: http://exbert.net Paper: https://arxiv.org/abs/1910.05276 #NLProc With @sebgehr @hen_str @MITIBMLab https://t.co/k4NrAgbCb7

9 replies, 707 likes


Sebastian Raschka: "ExBert - A Visual Analysis Tool to Explore Learned Representations in Transformers Models" -- this looks like a nice one! Tool: http://exbert.net and paper: https://arxiv.org/abs/1910.05276 https://t.co/q96JWrWfD2

5 replies, 326 likes


Suzana Ilić: A new interactive exploration tool. 😍

0 replies, 161 likes


Sasha Rush: New analysis tool from folks at MIT/IBM and Harvard NLP.

0 replies, 96 likes


Hendrik Strobelt: Tadaa 🎉 very happy to announce our newest project.. have fun exploring and playing with a Transfomer model.

0 replies, 50 likes


Sebastian Gehrmann: You liked all the analysis papers on Transformer attention? Now you can do your own experiments, with our exBERT framework!

3 replies, 41 likes


Jonathan Fly 👾: Lucid is great for images, but what about visualizing language models? exBERT: A Visual Analysis Tool to Explore Learned Representations in Transformers Models abs: https://arxiv.org/abs/1910.05276 Live Demo: http://exbert.net https://t.co/RvJ9zobPfT

0 replies, 31 likes


Tathagata: The @NeurIPSConf demo sessions are on! Come find @Ben_Hoov @_aabhas from @IBMResearch @IBM with cool tools for explainable #AI and deep learning. #NeurIPS2019 https://t.co/SVpHTvsva6

0 replies, 13 likes


Hendrik Strobelt: Join us for an ExBERT demo at #acl2020nlp on Tuesday: 1) 13:30-14:30 GMT / 09:30-10:30 EDT 2) 20:45-21:45 GMT / 16:45-17:45 EDT session: https://virtual.acl2020.org/paper_demo.84.html with @Ben_Hoov and @sebgehr #NLProc

0 replies, 10 likes


Lisa Amini: Cool visual analysis tool for BERT language models from our @IBMResearch Cambridge lab! It provides insights into learned contextual representations to help explain what each attention-head has learned.

0 replies, 4 likes


arXiv CS-CL: exBERT: A Visual Analysis Tool to Explore Learned Representations in Transformers Models http://arxiv.org/abs/1910.05276

0 replies, 2 likes


Mauro Martino: Another great project from our Lab!!

0 replies, 1 likes


午後のarXiv: "exBERT: A Visual Analysis Tool to Explore Learned Representations in Transformers Models", Benjamin Hoover, Hendri… https://arxiv.org/abs/1910.05276

0 replies, 1 likes


Brundage Bot: exBERT: A Visual Analysis Tool to Explore Learned Representations in Transformers Models. Benjamin Hoover, Hendrik Strobelt, and Sebastian Gehrmann http://arxiv.org/abs/1910.05276

1 replies, 0 likes


Content

Found on Oct 14 2019 at https://arxiv.org/pdf/1910.05276.pdf

PDF content of a computer science paper: EXBERT: A Visual Analysis Tool to Explore Learned Representations in Transformers Models