Search

There is a growing interest in integrating Large Language Models (LLMs) with autonomous driving (AD) systems. However, AD systems are vulnerable to attacks against their object detection and tracking (ODT) functions. Unfortunately, our evaluation of four recent LLM agents against ODT attacks shows that the attacks are 63.26% successful in causing them to crash or violate traffic rules due to (1) misleading memory modules that provide past experiences for decision making, (2) limitations of prompts in identifying inconsistencies, and (3) reliance on ground truth perception data. In this paper, we introduce Hudson, a driving reasoning agent that extends prior LLM-based driving systems to enable safer decision making during perception attacks while maintaining effectiveness under benign conditions. Hudson achieves this by first instrumenting the AD software to collect real-time perception results and contextual information from the driving scene. This data is then formalized into a domain-specific language (DSL). To guide the LLM in detecting and making safe control decisions during ODT attacks, Hudson translates the DSL into natural language, along with a list of custom attack detection instructions. Following query execution, Hudson analyzes the LLM's control decision to understand its causal reasoning process. We evaluate the effectiveness of Hudson using a proprietary LLM (GPT-4) and two open-source LLMs (Llama and Gemma) in various adversarial driving scenarios. GPT-4, Llama, and Gemma achieve, on average, an attack detection accuracy of 83. 3%, 63. 6%, and 73. 6%. Consequently, they make safe control decisions in 86.4%, 73.9%, and 80% of the attacks. Our results, following the growing interest in integrating LLMs into AD systems, highlight the strengths of LLMs and their potential to detect and mitigate ODT attacks.

High-level robot skills represent an increasingly popular paradigm in robot programming. However, configuring the skills' parameters for a specific task remains a manual and time-consuming endeavor. Existing approaches for learning or optimizing these parameters often require numerous real-world executions or do not work in dynamic environments. To address these challenges, we propose MuTT, a novel encoder-decoder transformer architecture designed to predict environment-aware executions of robot skills by integrating vision, trajectory, and robot skill parameters. Notably, we pioneer the fusion of vision and trajectory, introducing a novel trajectory projection. Furthermore, we illustrate MuTT's efficacy as a predictor when combined with a model-based robot skill optimizer. This approach facilitates the optimization of robot skill parameters for the current environment, without the need for real-world executions during optimization. Designed for compatibility with any representation of robot skills, MuTT demonstrates its versatility across three comprehensive experiments, showcasing superior performance across two different skill representations.

We numerically investigate the existence and stability dynamics of self-steepening optical solitons in a periodic PT-symmetric potential. We show that self-steepening solitons of the modified nonlinear Schr\"odinger (MNLS) equation undergo a position shift and amplitude increase during their evolution in the MNLS equation. The stabilization of solitons by an external potential is a challenging issue. This study demonstrates that the suppression of both the amplitude increase and the position shift of self-steepening solitons can be achieved by adding a periodic PT-symmetric potential to the MNLS equation.

This work introduces Variational Diffusion Distillation (VDD), a novel method that distills denoising diffusion policies into Mixtures of Experts (MoE) through variational inference. Diffusion Models are the current state-of-the-art in generative modeling due to their exceptional ability to accurately learn and represent complex, multi-modal distributions. This ability allows Diffusion Models to replicate the inherent diversity in human behavior, making them the preferred models in behavior learning such as Learning from Human Demonstrations (LfD). However, diffusion models come with some drawbacks, including the intractability of likelihoods and long inference times due to their iterative sampling process. The inference times, in particular, pose a significant challenge to real-time applications such as robot control. In contrast, MoEs effectively address the aforementioned issues while retaining the ability to represent complex distributions but are notoriously difficult to train. VDD is the first method that distills pre-trained diffusion models into MoE models, and hence, combines the expressiveness of Diffusion Models with the benefits of Mixture Models. Specifically, VDD leverages a decompositional upper bound of the variational objective that allows the training of each expert separately, resulting in a robust optimization scheme for MoEs. VDD demonstrates across nine complex behavior learning tasks, that it is able to: i) accurately distill complex distributions learned by the diffusion model, ii) outperform existing state-of-the-art distillation methods, and iii) surpass conventional methods for training MoE.

This work introduces Mamba Imitation Learning (MaIL), a novel imitation learning (IL) architecture that offers a computationally efficient alternative to state-of-the-art (SoTA) Transformer policies. Transformer-based policies have achieved remarkable results due to their ability in handling human-recorded data with inherently non-Markovian behavior. However, their high performance comes with the drawback of large models that complicate effective training. While state space models (SSMs) have been known for their efficiency, they were not able to match the performance of Transformers. Mamba significantly improves the performance of SSMs and rivals against Transformers, positioning it as an appealing alternative for IL policies. MaIL leverages Mamba as a backbone and introduces a formalism that allows using Mamba in the encoder-decoder structure. This formalism makes it a versatile architecture that can be used as a standalone policy or as part of a more advanced architecture, such as a diffuser in the diffusion process. Extensive evaluations on the LIBERO IL benchmark and three real robot experiments show that MaIL: i) outperforms Transformers in all LIBERO tasks, ii) achieves good performance even with small datasets, iii) is able to effectively process multi-modal sensory inputs, iv) is more robust to input noise compared to Transformers.

Integrated sensing and communications (ISAC) has emerged as a means to efficiently utilize spectrum and thereby save cost and power. At the higher end of the spectrum, ISAC systems operate at wideband using large antenna arrays to meet the stringent demands for high-resolution sensing and enhanced communications capacity. However, the wideband implementation entails beam-squint, that is, deviations in the generated beam directions because of the narrowband assumption in the analog components. This causes significant degradation in the communications capacity, target detection, and parameter estimation. This article presents the design challenges caused by beam-squint and its mitigation in ISAC systems. In this context, we also discuss several ISAC design perspectives including far-/near-field beamforming, channel/direction estimation, sparse array design, and index modulation. There are also several research opportunities in waveform design, beam training, and array processing to adequately address beam-squint in ISAC., Comment: Accepted Paper in IEEE Communications Magazine

In conventional remote sensing change detection (RS CD) procedures, extensive manual labeling for bi-temporal images is first required to maintain the performance of subsequent fully supervised training. However, pixel-level labeling for CD tasks is very complex and time-consuming. In this paper, we explore a novel self-supervised contrastive framework applicable to the RS CD task, which promotes the model to accurately capture spatial, structural, and semantic information through domain adapter and hierarchical contrastive head. The proposed SSLChange framework accomplishes self-learning only by taking a single-temporal sample and can be flexibly transferred to main-stream CD baselines. With self-supervised contrastive learning, feature representation pre-training can be performed directly based on the original data even without labeling. After a certain amount of labels are subsequently obtained, the pre-trained features will be aligned with the labels for fully supervised fine-tuning. Without introducing any additional data or labels, the performance of downstream baselines will experience a significant enhancement. Experimental results on 2 entire datasets and 6 diluted datasets show that our proposed SSLChange improves the performance and stability of CD baseline in data-limited situations. The code of SSLChange will be released at \url{https://github.com/MarsZhaoYT/SSLChange}, Comment: This manuscript has been submitted to IEEE TGRS and is under review

Interference Gating or iGate is a unique method for ultrafast time-resolved electron holography in a transmission electron microscope enabling a spatiotemporal resolution in the nm and ns regime with a minimal technological effort. Here, iGate is used for the first image-based investigation of the local dynamics of the projected electric potential in the area of the space charge region of two different general purpose silicon diodes during switching between unbiased and reverse biased condition with a temporal resolution of 25ns at a repetition rate of 3MHz. The obtained results for a focus-ion-beam-prepared ultrafast UG1A rectifier diode, which shows a decreasing capacitance with increasing reverse bias are in good agreement with an electric characterization of the macroscopic device as well as with theoretical expectations. For a severely modified 1N4007 device, however, time-resolved electron holography revealed a MOSCAP-like behavior with a rising capacitance in the area of the space charge region during the switching into reverse biased condition. Remarkably, a different behavior, dominated by the effective capacitance of the electrical setup, can be observed in the vacuum region outside both devices within the same measurements, clearly showing the benefits of localized dynamic potentiometry.

Y Cam is classified as one of the oscillating Eclipsing Algol (oEA) systems, which feature a $\delta$ Scuti-type pulsating component alongside mass transfer phenomena. oEA systems are invaluable for probing the evolutionary processes and internal structures of binary components offering insights through their binary variations and oscillating. In this study, we conducted a comprehensive investigation of Y Cam utilizing high-quality photometric TESS data, and high-resolution ELODIE spectra. Through our analysis, we examined the radial velocity variation, performed binary modeling, and calculated the effective temperature values of binary components. The fundamental stellar parameters, such as mass and radius, were determined with an accuracy of $\sim$ $2-6$ %. Furthermore, we examined the orbital period variation to assess the amount of mass transfer using the available minima times of the system and three new minima times obtained from TESS light curves. Analyzing the pulsation structure of the system with the TESS data revealed the dominant pulsation period and amplitude of the pulsating component to be 0.066 d and 4.65 mmag, respectively. Notably, we observed frequency modulations with the orbital period's frequency, along with variations in the amplitude of the highest amplitude frequency across different orbital phases. Remarkably, the amplitude reaches its peak at phases 0.5 and 1. These findings indicate a candidate of a tidally tilted pulsator. Consequently, we investigated the evolutionary status of the binary components using MESA binary evolution models, determining the age of the system to be 3.28 $\pm$ 0.09 Gyr., Comment: Accepted for publication in PASJ

Facial Expression Recognition (FER) plays a pivotal role in understanding human emotional cues. However, traditional FER methods based on visual information have some limitations, such as preprocessing, feature extraction, and multi-stage classification procedures. These not only increase computational complexity but also require a significant amount of computing resources. Considering Convolutional Neural Network (CNN)-based FER schemes frequently prove inadequate in identifying the deep, long-distance dependencies embedded within facial expression images, and the Transformer's inherent quadratic computational complexity, this paper presents the FER-YOLO-Mamba model, which integrates the principles of Mamba and YOLO technologies to facilitate efficient coordination in facial expression image recognition and localization. Within the FER-YOLO-Mamba model, we further devise a FER-YOLO-VSS dual-branch module, which combines the inherent strengths of convolutional layers in local feature extraction with the exceptional capability of State Space Models (SSMs) in revealing long-distance dependencies. To the best of our knowledge, this is the first Vision Mamba model designed for facial expression detection and classification. To evaluate the performance of the proposed FER-YOLO-Mamba model, we conducted experiments on two benchmark datasets, RAF-DB and SFEW. The experimental results indicate that the FER-YOLO-Mamba model achieved better results compared to other models. The code is available from https://github.com/SwjtuMa/FER-YOLO-Mamba.

Featurizing microscopy images for use in biological research remains a significant challenge, especially for large-scale experiments spanning millions of images. This work explores the scaling properties of weakly supervised classifiers and self-supervised masked autoencoders (MAEs) when training with increasingly larger model backbones and microscopy datasets. Our results show that ViT-based MAEs outperform weakly supervised classifiers on a variety of tasks, achieving as much as a 11.5% relative improvement when recalling known biological relationships curated from public databases. Additionally, we develop a new channel-agnostic MAE architecture (CA-MAE) that allows for inputting images of different numbers and orders of channels at inference time. We demonstrate that CA-MAEs effectively generalize by inferring and evaluating on a microscopy image dataset (JUMP-CP) generated under different experimental conditions with a different channel structure than our pretraining data (RPI-93M). Our findings motivate continued research into scaling self-supervised learning on microscopy data in order to create powerful foundation models of cellular biology that have the potential to catalyze advancements in drug discovery and beyond., Comment: CVPR 2024 Highlight. arXiv admin note: text overlap with arXiv:2309.16064

Large language models (LLMs) have become increasingly integrated with various applications. To ensure that LLMs do not generate unsafe responses, they are aligned with safeguards that specify what content is restricted. However, such alignment can be bypassed to produce prohibited content using a technique commonly referred to as jailbreak. Different systems have been proposed to perform the jailbreak automatically. These systems rely on evaluation methods to determine whether a jailbreak attempt is successful. However, our analysis reveals that current jailbreak evaluation methods have two limitations. (1) Their objectives lack clarity and do not align with the goal of identifying unsafe responses. (2) They oversimplify the jailbreak result as a binary outcome, successful or not. In this paper, we propose three metrics, safeguard violation, informativeness, and relative truthfulness, to evaluate language model jailbreak. Additionally, we demonstrate how these metrics correlate with the goal of different malicious actors. To compute these metrics, we introduce a multifaceted approach that extends the natural language generation evaluation method after preprocessing the response. We evaluate our metrics on a benchmark dataset produced from three malicious intent datasets and three jailbreak systems. The benchmark dataset is labeled by three annotators. We compare our multifaceted approach with three existing jailbreak evaluation methods. Experiments demonstrate that our multifaceted evaluation outperforms existing methods, with F1 scores improving on average by 17% compared to existing baselines. Our findings motivate the need to move away from the binary view of the jailbreak problem and incorporate a more comprehensive evaluation to ensure the safety of the language model.

Reinforcement learning (RL) is a powerful approach for acquiring a good-performing policy. However, learning diverse skills is challenging in RL due to the commonly used Gaussian policy parameterization. We propose \textbf{Di}verse \textbf{Skil}l \textbf{L}earning (Di-SkilL\footnote{Videos and code are available on the project webpage: \url{https://alrhub.github.io/di-skill-website/}}), an RL method for learning diverse skills using Mixture of Experts, where each expert formalizes a skill as a contextual motion primitive. Di-SkilL optimizes each expert and its associate context distribution to a maximum entropy objective that incentivizes learning diverse skills in similar contexts. The per-expert context distribution enables automatic curricula learning, allowing each expert to focus on its best-performing sub-region of the context space. To overcome hard discontinuities and multi-modalities without any prior knowledge of the environment's unknown context probability space, we leverage energy-based models to represent the per-expert context distributions and demonstrate how we can efficiently train them using the standard policy gradient objective. We show on challenging robot simulation tasks that Di-SkilL can learn diverse and performant skills., Comment: International conference on machine learning (ICML)

We report the discovery and characterisation of TIC 350842552 ("Zvrk"), an apparently isolated, rapidly-rotating ($P_\text{rot} \sim 99\ \mathrm{d}$) red giant observed by TESS in its Southern Continuous Viewing Zone. The star's fast surface rotation is independently verified by the use of p-mode asteroseismology, strong periodicity in TESS and ASAS-SN photometry, and measurements of spectroscopic rotational broadening. A two-component fit to APOGEE spectra indicates a coverage fraction of its surface features consistent with the amplitude of the photometric rotational signal. Variations in the amplitude of its photometric modulations over time suggest the evolution of its surface morphology, and therefore enhanced magnetic activity. We further develop and deploy new asteroseismic techniques to characterise radial differential rotation, and find weak evidence for rotational shear within Zvrk's convective envelope. This feature, in combination with such a high surface rotation rate, is incompatible with models of angular-momentum transport in single-star evolution. Spectroscopic abundance estimates also indicate a high lithium abundance, among other chemical anomalies. Taken together, all of these suggest a planet-ingestion scenario for the formation of this rotational configuration, various models for which we examine in detail., Comment: 31 pages, 17 figures. Accepted for publication in ApJ

The majority of data-driven wireless research leans heavily on discriminative AI (DAI) that requires vast real-world datasets. Unlike the DAI, Generative AI (GenAI) pertains to generative models (GMs) capable of discerning the underlying data distribution, patterns, and features of the input data. This makes GenAI a crucial asset in wireless domain wherein real-world data is often scarce, incomplete, costly to acquire, and hard to model or comprehend. With these appealing attributes, GenAI can replace or supplement DAI methods in various capacities. Accordingly, this combined tutorial-survey paper commences with preliminaries of 6G and wireless intelligence by outlining candidate 6G applications and services, presenting a taxonomy of state-of-the-art DAI models, exemplifying prominent DAI use cases, and elucidating the multifaceted ways through which GenAI enhances DAI. Subsequently, we present a tutorial on GMs by spotlighting seminal examples such as generative adversarial networks, variational autoencoders, flow-based GMs, diffusion-based GMs, generative transformers, large language models, to name a few. Contrary to the prevailing belief that GenAI is a nascent trend, our exhaustive review of approximately 120 technical papers demonstrates the scope of research across core wireless research areas, including physical layer design; network optimization, organization, and management; network traffic analytics; cross-layer network security; and localization & positioning. Furthermore, we outline the central role of GMs in pioneering areas of 6G network research, including semantic/THz/near-field communications, ISAC, extremely large antenna arrays, digital twins, AI-generated content services, mobile edge computing and edge AI, adversarial ML, and trustworthy AI. Lastly, we shed light on the multifarious challenges ahead, suggesting potential strategies and promising remedies.