Existing math datasets evaluate the reasoning abilities of large language models (LLMs) by either using the final answer or the intermediate reasoning steps derived from static examples. However, the former approach fails to surface model's uses of shortcuts and wrong reasoning while the later poses challenges in accommodating alternative solutions. In this work, we seek to use symbolic programs as a means for automated evaluation if a model can consistently produce correct final answers across various inputs to the program. We begin by extracting programs for popular math datasets (GSM8K and MATH) using GPT4-o. For those executable programs verified using the original input-output pairs, they are found to encapsulate the proper reasoning required to solve the original text questions. We then prompt GPT4-o to generate new questions using alternative input-output pairs based the extracted program. We apply the resulting datasets to evaluate a collection of LLMs. In our experiments, we observe significant accuracy drops using our proposed evaluation compared with original static examples, suggesting the fragility of math reasoning in state-of-the-art LLMs.

Maintaining an up-to-date map to reflect recent changes in the scene is very important, particularly in situations involving repeated traversals by a robot operating in an environment over an extended period. Undetected changes may cause a deterioration in map quality, leading to poor localization, inefficient operations, and lost robots. Volumetric methods, such as truncated signed distance functions (TSDFs), have quickly gained traction due to their real-time production of a dense and detailed map, though map updating in scenes that change over time remains a challenge. We propose a framework that introduces a novel probabilistic object state representation to track object pose changes in semi-static scenes. The representation jointly models a stationarity score and a TSDF change measure for each object. A Bayesian update rule that incorporates both geometric and semantic information is derived to achieve consistent online map maintenance. To extensively evaluate our approach alongside the state-of-the-art, we release a novel real-world dataset in a warehouse environment. We also evaluate on the public ToyCar dataset. Our method outperforms state-of-the-art methods on the reconstruction quality of semi-static environments.

CES may be full of robot's but not many have the same grip on humanity as those made by BrainCo. Among the towering TVs, smart appliances, and myriad other consumer-facing gadgets, BrainCo's line of fine-tuned prosthetic hands seem to reach out through the crowd - not just for their surprising capabilities, but for their ability to change people's lives. The hand is designed for amputees who have lost their forearm and is controlled by assessing the motion of the wearer's own muscles. While other prosthetics have achieved similar feats, BrainCo is granular enough that a wearer can move each and every individual finger, restoring not just grip but many fine-motor tasks. BrainCo's robotic hand (pictured above) could help amputees regain abilities to move fingers and grip items by reading electrical signals sent from the brain In a demonstration, the hand was able to pick up a cup and place it back down on the table.

The dream of self-driving cars is slowly becoming a reality. Best of all, you can take part of this revolution – at least on a small scale. Watch this talk by Ryan Zotti on how he converted a a toy RC car and made it autonomous.

So when you talk to us on the iPad, your audio is taken off the device through the cloud,

Google's self-driving car team is hiring executives as it prepares to spin out from Alphabet Audi's New AI Rig for Driverless Cars Is … a Toy Car Tourism to Boost the Self-Drive Car Rental Market in Emerging Countries Through 2020, Says ... Cadillac's Super Cruise Is Coming - Here's Why You'll Love It! Stay up-to-date on the topics you care about. We'll send you an email alert whenever a news article matches your alert term. It's free, and you can add new alerts at any time.

With that, Boris Sofman pulled a diminutive white toy out of a soft lunchbox, and began showing it off. After a brief introduction, Sofman set the little robot into its charging cradle, at which point it came to life in the way that all living beings recharge: It slept. I could tell Cozmo was sleeping because of the audio that was emanating from the speaker on his head: the sound of snoring. In between the purrs of his breaths, I heard a few seconds of music on what sounded like clarinet and strings. It was an expectant melody, a signal that Cozmo was ready to greet the day.