The new Kimi K3 leverages LatentMoE, a technique developed by @nvidia in January What is LatentMoE? In LatentMoE, tokens are projected from the model hidden dimension 𝑑 into a smaller latent dimension ℓ for expert routing and computation, which reduces routed parameter loads and all-to-all traffic by a factor of 𝑑/ℓ. Learn more about it here:

This is a really cool paper, sadly not open-source It sets a new SOTA on KITTI (relative depth estimation) and iBims-1 (metric depth estimation) Explore the paper and evals here:

As America turns 250, we put together 250 open AI milestones from the US: open models, datasets, demos, papers, and tools that helped shape the field. They go from attention is all you need, pytorch, gpt2, ULMFIT, llama, imagenet, Lora and hundreds more. They are a reminder of what made America the world’s engine of innovation: - Open science - Open competition - Open ecosystems Builders and scientists building on each other’s work, challenging each other, remixing ideas, and pushing the frontier forward. That is America at its best! And that philosophy is at risk right now in AI. In the coming months, scientists and AI builders will have to decide what side of history they want to be on: an AI future shaped by openness, transparency, participation, and competition, or one increasingly controlled behind closed doors by a few actors optimizing for money, secrecy, and gatekeeping. Let’s make the next 250 even more open!

Very proud to have spoken at @aiDotEngineer! Talked about automating my job at @huggingface with agents 🥷 Involves: > Claude Agents SDK > GLM-5.2 via Inference Providers > @langfuse for tracing > @modal for deployment Will be available on @YouTube later

Just added Lilian Weng's blog as a recommended read to the Scaling Laws method on Papers with Code Find the original paper + all papers which cite it, here:

Introducing a revival of PapersWithCode! As @ilyasut said, we're back to the "age of research". Hence, it's important to share research and build on each other's work. > find SOTA per domain, not just LLMs > leaderboards > methods > all parsed at scale using AI agents.