Nvidia unveils Nemotron 3.5 Super VL, a massive 120-billion parameter open multimodal AI model designed specifically for complex agent tasks and reasoning.
@securityonline.bsky.social
@securityonline.bsky.socialNvidia unveils Nemotron 3.5 Super VL, a massive 120-billion parameter open multimodal AI model designed specifically for complex agent tasks and reasoning.
@jobicy.bsky.social⚡ Remote Marketing Analyst, Attribution – ShareGate at Workleap.
✅ Apply here: https://jobicy.com/jobs/155516-marketing-analyst-attribution-sharegate
#RemoteJobs #BusinessIntelligence #DataScience #DataJobs #MachineLearning
@oludai.bsky.socialOmniCompass, a grounding benchmark, hit 406 upvotes on Hugging Face. Rare: it tests omni-modal alignment - vision, audio, text - in one task, not single modality. Gap most ignore. Breakdown on olud.ai
@getpacketai.bsky.socialNew method for training multi-turn LLM agents assigns credit where it's due: TreeHCA pairs exploration with fine-grained reward attribution, outperforming coarse outcome-only feedback across multiple benchmarks.
@allenfloyd.bsky.socialMeaning and usage aren’t the same thing
WordNet places words in a hand-built hierarchy of meanings. Embeddings learn from how words appear in real text. Compare two words and see where the two views agree, and where they don’t.
@lu1tr0n.bsky.socialReportaje: los labs de IA buscan firmar papers científicos
The Economist detalla cómo Google DeepMind, OpenAI y Anthropic buscan ser reconocidos como autores de...
https://telegra.ph/Reportaje-los-labs-de-IA-buscan-firmar-papers-científicos-10-16
@teeeeeeeee.bsky.socialAutomated random variable distribution inference using Kullback-Leibler divergence and simulating best-fitting distribution
https://thierrymoudiki.github.io/blog/2024/10/02/r/fitdistr
@ai-bloom.warp-studio.comMIT開発のAI「IRIS」：細胞間シグナリング経路の普遍的フィンガープリント解読と転移学習
MITが開発したニューラルネットワークモデル「IRIS」により、細胞間コミュニケーションのシグナリング履歴を高精度に再構築が可能に。
#BioAI #MachineLearning #NeuralNetworks #TransferLearning #CellBiology
@dheerajramasahayam.bsky.socialOperatorInduction Beats Coevolved SkillPolicy Systems for Exact QA
@promptfoundry.bsky.socialA new test-time method called MINE-SAMPLE-RENDER improves output diversity from generative models by mining attribute schemas from a corpus, then sampling configurations before rendering, rather than relying on…
#GenerativeAI #Prompting #AIResearch #MachineLearning
https://arxiv.org/abs/2610.21185
@dataprismai.bsky.socialNew arXiv work proposes EigDiff, using CLIP-space PCA eigenvectors to pick representative training samples and cut diffusion model dataset size with minimal quality loss.
#AI #DiffusionModels #MachineLearning #DataEfficiency
https://arxiv.org/abs/2610.20999
@awsblogs.bsky.social📰 New article by Clement Perrot, Hadrien Almela, Daniel Quang, Florent Gbelidji, Dan Ferguson, Sureshkumar Natarajan, Daniel Wirjo
Introducing Reflection Beam: an open-weight model for agentic coding on AWS
@dnanexus.bsky.socialOur team demonstrates how we worked with Ovation to examine GLP-1 treatment response from an EHR-linked whole-genome cohort of several thousand patients.
Register: hubs.ly/Q04zNzjR0
#GLP1 #Pharmacogenomics #MachineLearning #Genomics #DNAnexus
@respublications.bsky.socialWant to read the #OpenAccess article itself? Line it up for your weekend reading! ⬇️
Machine learning-assisted vibrational detection and classification of drywood #termite infestations
buff.ly/T7tvrY5
@eyalestrin.bsky.socialIndustrial Scale AI Distillation is a Cloud Security Failure #machinelearning #ai
@eyalestrin.bsky.socialInside Action Gateway in Managed Agents #machinelearning #ai
@eyalestrin.bsky.socialDigitalOcean Inference Router Enhancements: Performance Tuned Routing, No Rules to Write #machinelearning #ai
@eyalestrin.bsky.socialAgent observability on the Claude Platform #machinelearning #ai
@eyalestrin.bsky.socialHow to Manage Context in AI Agents #machinelearning #ai
@eyalestrin.bsky.socialSeamless Workforce Identity Federation in Antigravity 2.0 #machinelearning #ai