I am excited to announce a new version of deep-significance, version 1.2.6! 🥳 Below I'll talk about changes that were made while writing a demo paper, which can be found here: arxiv.org/pdf/2204.06815…, the package itself is at github.com/Kaleidophon/de… 🧵 (1/5)

I'm so happy to finally announce our work with Jes Frellsen and Christian Hardmeier at EMNLP 2025: We create a large benchmark of uncertainty quantification for #NLProc (8 models, up to 7 metrics) on three different languages! 🧵 (1/10) arxiv.org/abs/2210.15452 #emnlp2022

"Experimental Standards for Deep Learning in Natural Language Processing Research", an updated version of our workshop paper, got accepted at EMNLP 2025! #emnlp2022 🥳 Check the version below to learn how to make your empirical research more rigorous! arxiv.org/abs/2204.06251