Sd Deysangita·Oct 29From NLP Notes to a Story Worth Sharing — Part 1As a research analyst my first questions were about cells, genes and the science of life. I didn’t expect to find myself asking a different…
Johana Mayerova·Oct 12DaCy v2.9.0: Efficient models and simpler dependenciesRelease of improved small, medium and large DaCy models, and two new models, tiny and edge
Mohd Mustaqiim Asmuji (CHE QIIM)·Aug 22Menyanggah Label ‘Pseudosains’: Bukti Empirikal UiTM Buka Mata Isu NLP di MalaysiaAdakah Neuro-Linguistic Programming (NLP) sekadar ‘hiasan saintifik’ yang tidak mempunyai asas empirikal yang kukuh? Naratif yang sering…
Thet Twe Aung·Aug 10How I Converted 330K Burmese Words into 4.5 Million Romanized Spellings for a Pinyin-Style IMEBurglish is the Burmese input method: the user types romanized Burmese by ear (nay kaung lar), and the engine returns ranked Burmese…
Dr Krishna Athal·Aug 6Stop Generalising: 3 Ways to Speak Without AssumptionsYesterday evening, in the lift of a Mumbai office tower, I heard one young professional tell another, “HR people are all the same, they…
InSupportivbySiddharth Shukla·Jul 28Part 4: Downstream Bias Propagation — Tracing Pre-Training Choices Through Fine-Tuned ModelsSharing these results, including the nuances and tradeoffs, contributes to the understanding of how bias propagates through the ML pipeline.A response icon1A response icon1
Sevvalmikci·Jul 19LLM (Büyük Dil Modelleri) Nedir? Yapay Zekanın “Beyni” Nasıl Çalışıyor?Yapay Zekanın “Sessiz Okuyucusu”
Simon P·Jul 7Warning: Doing Affirmations Without This One Step Is Worse Than Doing Nothing.The Invisible Saboteur Living Rent-Free in Your Head
InSupportivbySiddharth Shukla·Jun 30Part 2: Pre-training RoBERTa from Scratch with Curriculum Learning and Bias-Aware LossTraining a model to learn equitable representations across demographic groups while maintaining its strong language modeling ability.A response icon2A response icon2
