Decoding the future @ Litus AI ~ NLP PhD @ Sapienza - Engineer at heart - he/him.
Roma
Joined January 2012
- Tweets61
- Following126
- Followers106
- Likes103
Niccolò Campolungo retweeted
📢 It looks like relative representations are here to stay!
I'm beyond thrilled to announce that our work has been selected as one of the notable top 5% (oral) papers at #iclr23 ! 🥳
teapawt.org/moschella_luca/status/…
[1/5]
Welcome Relative Representations, enabling zero-shot communication between latent spaces without any training!
arxiv.org/abs/2209.15430
It turns out that distinct neural networks learn intrinsically equivalent latent spaces [1/6]
Niccolò Campolungo retweeted
We reframe #MT evaluation as sequence tagging in our #WMT2022 paper: check MaTESe, an interpretable metric that tags error spans in translations!
📜 statmt.org/wmt22/pdf/2022.wm…
By @19Stefano97 @LorenzoProiet13 @alescire94 @Valahaar @RNavigli @babelscape @ERC_Research #EMNLP2022
Niccolò Campolungo retweeted
Great to wake up on a lazy Sunday and see REBEL is still doing numbers. This is the page for top seq2seq models at @huggingface, currently top 12 with 200K monthly downloads. Thank you to whomever is using it and honestly curious to know what people are using it for.
Niccolò Campolungo retweeted
Interested in reducing semantic bias in NMT models with smart data creation & the KL divergence?
Come check out the work by @Valahaar, @pasini_t, @emelin_denis & @RNavigli!
Presentation today @ 9.15 PDT / 18.15 CET, session 8D (MT 3) #NAACL2022
📝 aclanthology.org/2022.naacl-…
Niccolò Campolungo retweeted
machine learning researchers learn to optimise their own best paper rate through collusion and other unregulated mechanisms
reddit.com/r/MachineLearning…
Niccolò Campolungo retweeted
Hey #NLProc, I built this little tool to make working with @huggingface 🤗Transformers a bit easier. If you want to directly access whole-word embeddings hassle-free, give it a try!
👉GitHub: github.com/Riccorl/transform…
Niccolò Campolungo retweeted
The Rome Workshop on 10 Years of #BabelNet & Multilingual Neurosymbolic Natural Language Understanding was a great success, with productive in-person discussions, amazing talks & >100 online participants! Thanks!
@ERC_Research @Babelscape @SapienzaNLP @SapienzaRoma @WikiResearch
Niccolò Campolungo retweeted
🎉10 Years of #BabelNet Workshop is now over:
what a great experience!🎉
Thanks to everyone who participated both in person and online!
@SapienzaNLP @babelscape @ERC_Research
@AccademiaCrusca @RNavigli
#workshop @WikiResearch #BabelNet10anniversary
Niccolò Campolungo retweeted
🥳 We are proud to share the news that our DiBiMT paper on Disambiguation Biases in MT received the ✨Best Resource Paper Award✨ @aclmeeting #ACL2022! See you there!
Open & commercial Neural Machine Translation models heavily suffer from disambiguation biases! We present DiBiMT, our novel benchmark for lexical-semantic bias in MT at #ACL2022! By @Valahaar @FedeMartelli25 @FrancescoSaina @RNavigli
@ELEXIS_EU #NLProc
📝:researchgate.net/publication…
Niccolò Campolungo retweeted
So proud that my former labmates at @SapienzaNLP have been awarded the Best Resource Paper at the #ACL2022 for their work on benchmarking disambiguation in MT! Congrats @Valahaar @FedeMartelli25 @FrancescoSaina @RNavigli
The Best Paper Awards have just been announced! Congratulations to the authors for their fine work!
2022.aclweb.org/best-paper-a…
Niccolò Campolungo retweeted
New work on autoregressive language models for retrieval!
We train our model, SEAL (Search Engines with Autoregressive LMs) to produce text snippets occurring somewhere in the corpus, in relevant documents. We use the generated ngrams to rank documents in the corpus.
Thread 👇
Niccolò Campolungo retweeted
#NLPaperAlert 📢: where do language models encode predicate-argument structure information? And how to use it in multilingual Semantic Role Labeling?
Find out more in our #ACL2022 paper! #NLProc
📝: researchgate.net/publication…
By @ConiaSimone @RNavigli
+ @ERC_Research @EuroLangTech
Niccolò Campolungo retweeted
We analyzed current Word Sense Disambiguation #WSD systems so that you don’t have to 🔍. There is a hard core that no system can solve! 😱 #NLProc
Read our #ACL2022 paper: researchgate.net/publication…
By @MarcoMaru3 @ConiaSimone @MicheleBevila20 @RNavigli + @ERC_Research @elexis_eu
Niccolò Campolungo retweeted
"DIBIMT: A Novel Benchmark for Measuring Word Sense Disambiguation Biases in Machine Translation" a 5-language benchmark to study semantic biases in machine translation, based on @Wiktionary
(Campolungo et al, 2022)
data: nlp.uniroma1.it/dibimt
paper: researchgate.net/profile/Fed…
Niccolò Campolungo retweeted
Classy is a @PyTorch-based library for the fast prototyping and sharing of deep neural network models.
It wraps the best libraries like PyTorch Lightning, Transformers, @streamlit and offers them to users with a simple CLI interface.
Try it here: github.com/sunglasses-ai/cla…
Niccolò Campolungo retweeted
Extractive Entity Disambiguation is our (@edoardo_barba @luigi_proc @RNavigli) new paper at #ACL2022! Guess what, we perform ED as a Text Extraction task using only Wikipedia Titles and achieve a new SotA in standard and constrained data regimes.
@ERC_Research #NLProc
Links: ⬇️
Niccolò Campolungo retweeted
Brief update, we decided to give this a chance with some changed/new roles (ie. Reviewer 2 😈) and it has been very fun and showing promise. There's room for improvement but I can recommend to give it a try. Hopefully next time we'll do it offline 😃
Adding a non-Twitter contribution by @Valahaar: colinraffel.com/blog/role-pl…
Thanks for the other suggestions as well!
Niccolò Campolungo retweeted
#BabelNet is now an app! 20 million meanings and millions of definitions in 500 languages!
Google Play 👉 bit.ly/3uWZARk
Apple store 👉 apple.co/3rTSB9v
#applaunch #applaunchday #applaunching #AppStore #PlayStore @SapienzaNLP @babelscape
Niccolò Campolungo retweeted
Last week remote working with @edoardo_barba and @Valahaar: skiing in the Alps while checking runs with @wandb
