Rebecca Pattichis
22 posts

Rebecca Pattichis
@rpattichi
all things multilingual and code-switching NLP | Previously DeepMind Fellow MSCS @ UCLA, BSH @ Stanford




1/n @uclanlp is researching how Black, LGBTQIA+, & women communities perceive and are affected by content moderation, as it relates to English-language social media content using reclaimed language. As part of this, we are recruiting annotators (forms.gle/KP6F9gDCo8Skjs…) …


Trending in #Literature: ooir.org/index.php?fiel… 1) Nothing But the Fictions 2) Reading It Wrong: An Alternative History of Early 18th Century Literature [book review] (@English_Jnl) 3) How artists justify career paths (@poetics_journal) 4) Shakespeare’s Love’s Labour’s Lost in Margaret Cavendish’s The Convent of Pleasure 5) Multilingualism & mismatching: Spanish language usage in college admissions essays

Sharing a new paper that examines how often Spanish words/phrases were used in 275,000+ college admission essays. The short answer: quite a bit! Using some form of Spanish was actually a more common occurrence than references to parental divorce sciencedirect.com/science/articl…



🚨 Previous @Gemini incidence revealed severe fairness concerns in Text-To-Image (T2I) models, leading to increased efforts to research biases in T2I systems. At this critical time point, we take a retrospective step and reflect on ‘What has been done’ and ‘What can be done’ in this direction. Check out our latest work, the FIRST Survey on Bias in T2I Models! 🤩 We comprehensively review bias definition, evaluation, and mitigation methods in previous studies, point out current research gaps, and shed light on potential future steps to tackle biases in T2I systems. 📚 🔗Link to full paper: arxiv.org/abs/2404.01030

🚨Video Out! Check out our thoughts & ideas on "Working with Low-Resource Language Research/Applications": youtu.be/TlexQmpJtAE Thank you to our amazing panelists and hosts: @anas_ant @luischiruzzo @OanaIgnatRo @ZhijingJin


