“🚨This week’s top AI/ML research papers: – LLaVA-o1 – Marco-o1 – The Dawn of GUI Agent – Hymba – When Precision Meets Position – Multimodal Autoregressive Pre-training of Large Vision Encoders – Generative World Explorer – That Chip Has Sailed – Is Your LLM Secretly a World
Building an AI-Powered Game – DeepLearning.AI
Do Coding Boot Camps Make Sense in an A.I. World? – The New York Times
2024: The State of Generative AI in the Enterprise – Menlo Ventures
“@MattMcLx Isn’t torch.distributed all about just easing the hassle of managing multiple processes? I mean literally a single process.” / X
[2411.14343v1] UnifiedCrawl: Aggregated Common Crawl for Affordable Adaptation of LLMs on Low-Resource Languages
“I’m excited to also announce that we’ll be doing the first Latent Space LIVE! at @NeurIPSConf! Both online + IRL. For the first time, a full day side event with all the stuff I found missing from standard academic conferences: – Too Hot For NeurIPS (papers too new/rejected
Grounding-IQA
StableAnimator
LLM-as-a-judge
[2411.17116v1] Star Attention: Efficient LLM Inference over Long Sequences
[2411.15131] WildLMa: Long Horizon Loco-Manipulation in the Wild
blog – Flow With What You Know
“🚨 New Paper 🚨 Can LLMs perform latent multi-hop reasoning without exploiting shortcuts? We find the answer is yes – they can recall and compose facts not seen together in training or guessing the answer, but success greatly depends on the type of the bridge entity (80%+ for
“Abacus AI Introduces @github Automation Using AI To write Code, Fix code Using AI to create Pull Request from the same Chat Interface (ChatLLM) đź“š Key Features ⚙️ How to Install? 🚀 How to Use? đź’ˇ Key Benefits 🔄 Workflow Steps: @abacusai
“We’re hiring an intern to join our SmolLM team and help build the next generation of smol models. If you’re passionate about training LLMs and curating high-quality datasets, we’d love to hear from you!
“Excited for our Benchmarking Hub: independent evals to build shared understanding of AI capabilities. Coming soon: • More benchmarks: FrontierMath, SWE-Bench (basically all the best ones) • Predictive work—aiming to be for AI what FiveThirtyEight is for elections” / X
Gwern Branwen – How an Anonymous Researcher Predicted AI’s Trajectoryhttps://www.dwarkeshpatel.com/p/gwern-branwen
Leave a Reply