For friends of open source: imo the highest leverage thing you can do is help construct a high diversity of RL environments that help elicit LLM cognitive strategies. To build a gym of sorts. This is a highly parallelizable task, which favors a large community of collaborators.

We propose new scaling laws that predict the optimal data mixture, for pretraining LLMs, native multimodal models and large vision encoders ! Only running small-scale experiments is needed, and we can then extrapolate to large-scale ones. These laws allow 1/n 🧵

EarthCrafter Scalable 3D Earth Generation via Dual-Sparse Latent Diffusion

Everything about Llama-Nemotron-Super-V1.5 post-training is now open: Synthetic data: huggingface.co/datasets/nvidi… Human data: huggingface.co/datasets/nvidi… Reward models (trained on HS3 data): huggingface.co/collections/nv… RL toolkit: github.com/NVIDIA-NeMo/RL

Hierarchical Reasoning Model This is one of the most interesting ideas on reasoning I've read in the past couple of months. It uses a recurrent architecture for impressive hierarchical reasoning. Here are my notes:

What if you could not only watch a generated video, but explore it too? 🌐 Genie 3 is our groundbreaking world model that creates interactive, playable environments from a single text prompt. From photorealistic landscapes to fantasy realms, the possibilities are endless. 🧵

ICLR 2026 will take place in 📍Rio de Janeiro, Brazil 📅 April 23–27, 2026 Save the date - see you in Rio! #ICLR2026

Analyzing the Hierarchical Reasoning Model by Guan Wang We verified scores on hidden tasks, ran ablations, and found that performance comes from an unexpected source ARC-AGI Semi Private Scores: * ARC-AGI-1: 32% * ARC-AGI-2: 2% Our 4 findings:

We were able to reproduce the strong findings of the HRM paper on ARC-AGI-1. Further, we ran a series of ablation experiments to get to the bottom of what's behind it. Key findings: 1. The HRM model architecture itself (the centerpiece of the paper) is not an important factor.

Environments and agentic interactions🚀

x.com/i/article/1960…

My pleasure to come on Dwarkesh last week, I thought the questions and conversation were really good. I re-watched the pod just now too. First of all, yes I know, and I'm sorry that I speak so fast :). It's to my detriment because sometimes my speaking thread out-executes my

On-policy distillation provides an elegant way to use the teacher model as a process reward model to provide dense reward while preventing SFT style "OOD shock" during rollout.

🎃 This Halloween, Explore Multimodal Universe! 🚀 Explore, navigate, and interact across space and time — all with long-horizon consistency. Guess how many worlds Emu3.5 jumps through? 👀👀 #Emu3_5 #WorldModel #AI #Halloween2025

Classic scaling law curves from Generalist: more data, lower validation loss — standard ML story. But if we swap the y-axis to real-world success rate, do the same trends hold? Does 0.0105 vs 0.0115 loss actually mean a more reliable policy?

When we began applying diffusion to language in my lab at Stanford, many doubted it could work. That research became Mercury diffusion LLM: 10X faster, more efficient, and now the foundation of Inception. Proud to raise $50M with support from top investors.