10/25 𝗩𝗶𝗿𝗮𝗹 𝗣𝗿𝗼𝘁𝗲𝗶𝗻𝘀 𝗥𝗲𝘃𝗲𝗮𝗹 𝗚𝗲𝗼𝗺𝗲𝘁𝗿𝘆 𝗼𝗳 𝗣𝗿𝗼𝘁𝗲𝗶𝗻 𝗟𝗮𝗻𝗴𝘂𝗮𝗴𝗲 𝗠𝗼𝗱𝗲𝗹𝘀 This paper investigates how protein language models (pLMs) represent underrepresented viral proteins, identifying a dominant 'nativeness axis' in embedding space aligned with masked reconstruction perplexity. While this axis contracts unevenly across viral families, pLM embeddings retain viral-specific signal, demonstrating linear separability and suggesting representations balance a general nativeness concept with distinct biological group information. #ProteinLanguageModels #pLM #ESMModels #ViralProteins #RepresentationalLearning #BiologicalSequences Paper Link: arxiv.org/abs/2606.12609

How can #AI machines learn to see and understand visual information like the human brain? In his inaugural #lecture Prof @STsaftaris will explore #RepresentationalLearning and its applications, from restoring art to enhancing medical imaging diagnosis. edin.ac/3xuSt5G

Yann LeCun’s Deep Learning Course at CDS is Now Fully Online & Accessible to All buff.ly/3dhuwAw #ai #machinelearning #artificialintelligence #RepresentationalLearning #DeepLearning via @ylecun

Need a deep dive on #DeepLearning, #RepresentationalLearning, #computervision and more AI? Prof. @ylecun's #online 14-week course @NYUDataScience is free and accessible to all in 12 languages! bit.ly/36HHgiR

My team @VMware continues to innovate with 3 more patents filed spanning traditional application of #statisticallearning, #representationallearning and classic #machinelearning to #cybersecurity and #cyberhygiene. Happy Diwali !

Awesome and thought provoking talk by #YaelNiv from @Princeton about #reinforcementLearning & #representationalLearning