We asked attendees of Super Bowl LX's pregame festivities for their takes on the competing halftime shows, the potential for ICE actions, and the influence of Silicon Valley on the event. To say this year's Super Bowl came at a charged time in American culture and politics is, perhaps, an understatement. While the pair of teams who took the field Sunday--the Seattle Seahawks and the New England Patriots--comprised a pretty classic matchup (no underdogs here!), the rest of the event was set to be anything but. Santa Clara's Levi's Stadium is in the heart of Silicon Valley, just a few miles from the corporate headquarters of Nvidia and AMD, whose chips are powering the AI arms race that had competitors OpenAI and Anthropic sparring via rival Super Bowl ads . There was an explosion in sports "trading" activity on sites like Kalshi and Polymarket in the lead-up to the game, even in states like California where traditional sports betting is illegal. Sunday could prove to be an extraordinary success for prediction markets, as the industry becomes more mainstream . Fresh off a historic Grammy Album of the Year win (a first for a Spanish-language album), the unapologetically political Puerto Rican rapper and singer Bad Bunny headlined --a choice that sparked a perhaps inevitable MAGA backlash. Meanwhile, Turning Point USA organized an alternative program called The All-American Halftime Show, featuring the likes of Kid Rock and Brantley Gilbert. Never mind that Bad Bunny is Puerto Rican, and therefore an American citizen. Rumors were even buzzing about possible actions by US Immigration and Customs Enforcement agents at the Super Bowl. Even though the NFL and California governor Gavin Newsom said on Thursday that there would be " no immigration enforcement tied to the game," anti-ICE protesters were on the streets. We caught up with football fans at a tailgate five minutes away from Levi's Stadium to hear their thoughts on all the drama. Here's what they had to say.

As digitalization and cloud technologies evolve, the web is becoming increasingly important in the modern society. Autonomous web agents based on large language models (LLMs) hold a great potential in work automation. It is therefore important to accurately measure and monitor the progression of their capabilities. In this work, we conduct a comprehensive and rigorous assessment of the current state of web agents. Our results depict a very different picture of the competency of current agents, suggesting over-optimism in previously reported results. This gap can be attributed to shortcomings in existing benchmarks. We introduce Online-Mind2Web, an online evaluation benchmark consisting of 300 diverse and realistic tasks spanning 136 websites. It enables us to evaluate web agents under a setting that approximates how real users use these agents. To facilitate more scalable evaluation and development, we also develop a novel LLM-as-a-Judge automatic evaluation method and show that it can achieve around 85% agreement with human judgment, substantially higher than existing methods. Finally, we present the first comprehensive comparative analysis of current web agents, highlighting both their strengths and limitations to inspire future research.

Accurate and efficient surgical robotic tool pose estimation is of fundamental significance to downstream applications such as augmented reality (AR) in surgical training and learning-based autonomous manipulation. While significant advancements have been made in pose estimation for humans and animals, it is still a challenge in surgical robotics due to the scarcity of published data. The relatively large absolute error of the da Vinci end effector kinematics and arduous calibration procedure make calibrated kinematics data collection expensive. Driven by this limitation, we collected a dataset, dubbed SurgPose, providing instance-aware semantic keypoints and skeletons for visual surgical tool pose estimation and tracking. By marking keypoints using ultraviolet (UV) reactive paint, which is invisible under white light and fluorescent under UV light, we execute the same trajectory under different lighting conditions to collect raw videos and keypoint annotations, respectively. The SurgPose dataset consists of approximately 120k surgical instrument instances (80k for training and 40k for validation) of 6 categories. Each instrument instance is labeled with 7 semantic keypoints. Since the videos are collected in stereo pairs, the 2D pose can be lifted to 3D based on stereo-matching depth. In addition to releasing the dataset, we test a few baseline approaches to surgical instrument tracking to demonstrate the utility of SurgPose. More details can be found at surgpose.github.io.