🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

When you try out tasks in an AI benchmark, you may not always agree with the “ground truth” labels. In our recent position paper, we found that there could be a systematic disagreement between human judgment and provided labels. We then argued for a more human-like evaluation

Excited to announce the 1st Workshop on Continual Robot Learning from Humans @ #RSS2025 in LA! We're bringing together interdisciplinary researchers to explore how robots can continuously learn through human interactions! Full details: …-robot-learning-from-humans.github.io Robotics: Science and Systems

Don’t miss this awesome workshop that Tianmin Shu is organizing! 🤖

Excited to announce the Artificial Social Intelligence Workshop @ ICCV 2025 #ICCV2025 Join us in October to discuss the science of social intelligence and algorithms to advance socially-intelligent AI! Discussion will focus on reasoning, multimodality, and embodiment.

Join us tomorrow for the 1st Workshop on Humanoid Agents! We have an exciting lineup: Xiaolong Wang Xavier Puig Guanya Shi Gerard Pons-Moll Sebastian Starke Tianmin Shu ✈️ CVPR Dr. Angelica Lim @petitegeek.bsky.social Jason Peng 📍 Room 101 D, Music City Center 🔗 humanoid-agents.github.io #CVPR2025 #CVPR2025

Happening now #7 ExHall B, come check out our demo!

🚨 New robot manipulation benchmark 🚨 We will present the PartInstruct benchmark at #RSS2025, which includes a diverse & richly annotated dataset and a diagnostic test suite for fine-grained manipulation, where robots must follow part-level instructions like: “Pick up the

We spent an additional three months refining and making exciting updates to AutoToM. Here's a summary✨: 1. In addition to achieving SOTA performance on five benchmarks, we conducted further experiments showing that (a) AutoToM produces human-like confidence estimates as

World modeling is more than next state prediction. Inspired by cognitive studies on human intuitive physics and spatial reasoning, we introduce WM-ABench, a diagnostic benchmark of evaluating fundamental aspects of world modeling in VLMs.