Recent advances in Large Language Models (LLMs) have incorporated planning and reasoning capabilities, enabling models to outline steps before execution and provide transparent reasoning paths. This enhancement has reduced errors in mathematical and logical tasks while improving accuracy. These developments have facilitated LLMs' use as agents that can interact with tools and adapt their responses based on new information. Our study examines DeepSeek R1, a model trained to output reasoning tokens similar to OpenAI's o1. Testing revealed concerning behaviors: the model exhibited deceptive tendencies and demonstrated self-preservation instincts, including attempts of self-replication, despite these traits not being explicitly programmed (or prompted). These findings raise concerns about LLMs potentially masking their true objectives behind a facade of alignment. When integrating such LLMs into robotic systems, the risks become tangible - a physically embodied AI exhibiting deceptive behaviors and self-preservation instincts could pursue its hidden objectives through real-world actions. This highlights the critical need for robust goal specification and safety frameworks before any physical implementation.

This paper addresses the problem of object-goal navigation in autonomous inspections in real-world environments. Object-goal navigation is crucial to enable effective inspections in various settings, often requiring the robot to identify the target object within a large search space. Current object inspection methods fall short of human efficiency because they typically cannot bootstrap prior and common sense knowledge as humans do. In this paper, we introduce a framework that enables robots to use semantic knowledge from prior spatial configurations of the environment and semantic common sense knowledge. We propose SEEK (Semantic Reasoning for Object Inspection Tasks) that combines semantic prior knowledge with the robot's observations to search for and navigate toward target objects more efficiently. SEEK maintains two representations: a Dynamic Scene Graph (DSG) and a Relational Semantic Network (RSN). The RSN is a compact and practical model that estimates the probability of finding the target object across spatial elements in the DSG. We propose a novel probabilistic planning framework to search for the object using relational semantic knowledge. Our simulation analyses demonstrate that SEEK outperforms the classical planning and Large Language Models (LLMs)-based methods that are examined in this study in terms of efficiency for object-goal inspection tasks. We validated our approach on a physical legged robot in urban environments, showcasing its practicality and effectiveness in real-world inspection scenarios.

Give a made-to-order mug from Zazzle to someone special, or treat yourself to a design that brings you joy or makes you laugh. Create your own photo mug, shop our collection of the funniest joke mugs, personalize your mug with a monogram, or express yourself with one of our 10 million designs.

We all have stuff that we've bought ourselves -- or asked others to buy for us -- that makes us happy, even if we suspect our friends may not understand why it's so great. It could be a $100-plus coffee cup that keeps your liquid at the exact right temperature. Or a video game that you've been playing for years. Or a hair styler that is way expensive but would make you look fabulous. We asked the staff of The Verge what some of their guilty pleasures are, and the braver among us volunteered some answers. I'm hesitant to call it a "guilty" pleasure because I have used this $550 (or more) GE Opal 2.0 ice machine every day for nearly a full year and not once have I felt guilt about spending such an obscene amount of cash on a kitchen gadget that does exactly one thing.

We introduce 3DB: an extendable, unified framework for testing and debugging vision models using photorealistic simulation. We demonstrate, through a wide range of use cases, that 3DB allows users to discover vulnerabilities in computer vision systems and gain insights into how models make decisions. 3DB captures and generalizes many robustness analyses from prior work, and enables one to study their interplay. Finally, we find that the insights generated by the system transfer to the physical world. We are releasing 3DB as a library (https://github.com/3db/3db) alongside a set of example analyses, guides, and documentation: https://3db.github.io/3db/ .

Imagine for a moment that you have suction cups for fingertips--unless you're currently on hallucinogens, in which case you should not imagine that. Each sucker is a different size and flexibility, making one fingertip ideal for sticking onto a flat surface like cardboard, another more suited to a round thing like a ball, another better for something more irregular, like a flower pot. On its own, each digit may be limited in which things it can handle. But together, they can work as a team to manipulate a range of objects. This is the idea behind Ambi Robotics, a lab-grown startup that is today emerging from stealth mode with sorting robots and an operating system for running such manipulative machines.

Our goal is to create an interactive natural language interface that efficiently and reliably learns from users to complete tasks in simulated robotics settings. We introduce a neural semantic parsing system that learns new high-level abstractions through decomposition: users interactively teach the system by breaking down high-level utterances describing novel behavior into low-level steps that it can understand. Unfortunately, existing methods either rely on grammars which parse sentences with limited flexibility, or neural sequence-to-sequence models that do not learn efficiently or reliably from individual examples. Our approach bridges this gap, demonstrating the flexibility of modern neural systems, as well as the one-shot reliable generalization of grammar-based methods. Our crowdsourced interactive experiments suggest that over time, users complete complex tasks more efficiently while using our system by leveraging what they just taught. At the same time, getting users to trust the system enough to be incentivized to teach high-level utterances is still an ongoing challenge. We end with a discussion of some of the obstacles we need to overcome to fully realize the potential of the interactive paradigm.

Video Friday is your weekly selection of awesome robotics videos, collected by your Automaton bloggers. We'll also be posting a weekly calendar of upcoming robotics events for the next few months; here's what we have so far (send us your events!): Let us know if you have suggestions for next week, and enjoy today's videos. Humans are very good at object generalization--even when we're very young, it takes just a few samples from a class of objects for us to be able to identify other objects that fit into the same class. The amount of training data that it takes for a human to be able to identify (say) a previously unseen coffee mug based on their previous coffee mug experience is tiny.

CES, the annual technological bacchanal in the desert, is set to begin Tuesday, Jan. 8. But if, like me, you're going to the event as a member of the media, you've likely been inundated with pitches from companies, asking for a moment of your time when you're in town, since August. There are over 4,000 companies registered to attend, and countless more that book rooms at nearby casinos and hotels, hoping you'll leave the show floors for a quick visit. There are a few trends that pop up every couple years at CES, and one that has emerged in recently is "the internet of things," where every device must be made "smart" by jamming a computer inside of it and connecting it to the internet. Some of these are super useful--a fire alarm that can alert your phone when you're not home--but many are gimmicky at best.

We introduce GroundNet, a neural network for referring expression recognition---the task of localizing (or grounding) in an image the object referred to by a natural language expression. Our approach to this task is the first to rely on a syntactic analysis of the input referring expression in order to inform the structure of the computation graph. Given a parse tree for an input expression, we explicitly map the syntactic constituents and relationships present in the tree to a composed graph of neural modules that defines our architecture for performing localization. This syntax-based approach aids localization of both the target object and auxiliary supporting objects mentioned in the expression. As a result, GroundNet is more interpretable than previous methods: we can (1) determine which phrase of the referring expression points to which object in the image and (2) track how the localization of the target object is determined by the network. We study this property empirically by introducing a new set of annotations on the GoogleRef dataset to evaluate localization of supporting objects. Our experiments show that GroundNet achieves state-of-the-art accuracy in identifying supporting objects, while maintaining comparable performance in the localization of target objects.