🎉📢Exciting news! Join us at the inaugural #CVPR2025 workshop on 3D Vision Language Models (VLMs) for Robotic Manipulation: Opportunities and Challenges, happening on June 11, 2025, in Nashville, TN. Explore how 3D perception can be integrated into robotic manipulation in the

happy 8th birthday of maybe the most influential paper of all time 🥳

Heya roboticists + mind minders, I put together a list of 100+ robotics papers from CVPR this year - with TLDRs, virtual posters + interactive demos on Bytez (we're proud to be embedded in the conference) CVPR vision models coming soon! github.com/smallfryy/cvpr… #CVPR2025

#CVPR2025 paper stats

Thrilled and honored to receive the Best Paper Award at #CVPR2025! Huge thanks to my fantastic collaborators Minghao Chen, Nikita Karaev, Andrea Vedaldi, Christian Rupprecht, and David Novotny. Could not be there without you!

Congrats to Jianyuan#CVPR2025 2025 Minghao Chen Nikita Karaev, Andrea Vedaldi, Christian Rupprecht & David Novotny for winning the Best Paper Award at @CVPR for "VGGT: Visual Geometry Grounded Transformer" 🎉 Explore, listen to & get code for their paper here: bytez.com/docs/cvpr/3396…

This weekend took place the worldwide LeRobot hackathon, with over 1,000 mind-blowing demos from 100 different local editions. So, I listed the ones I liked the most from robots killing bugs, using chopticks or playing air hockey. Here they are:

Golly gee, they like us! Turns out Llamas and Oxen go together like chocolate and vanilla! We're excited to announce we're a part of the first Llama Startup Program. If you're looking to easily fine-tune any Llama models on your data - hit us up. 🦙🐂

Introducing Reinforcement-Learned Teachers (RLTs): Transforming how we teach LLMs to reason with reinforcement learning (RL). Blog: sakana.ai/rlt Paper: arxiv.org/abs/2506.08388 Traditional RL focuses on “learning to solve” challenging problems with expensive LLMs and

How novel is your research? Meet KnoVo, an LLM tool to measure research novelty by comparing a paper’s contributions to others in its citation network Extracts key dimensions (technique, dataset, etc) Scores novelty across those dimensions Visualizes how ideas evolve over time

finally got around to fixing the chat template for the rest of the Qwen3 models

Open-source superintelligence for all!

Introducing SciArena, a platform for benchmarking models across scientific literature tasks. Inspired by Chatbot Arena, SciArena applies a crowdsourced LLM evaluation approach to the scientific domain. 🧵

Are you headed to ICML 2025 in Vancouver? Register for our happy hour with @AskAlphaXiv!

Introducing IFBench, a benchmark to measure how well AI models follow new, challenging, and diverse verifiable instructions. Top models like Gemini 2.5 Pro or Claude 4 Sonnet are only able to score up to 50%, presenting an open frontier for post-training. 🧵