Despite the general capabilities of large pretrained language models, they consistently benefit from further adaptation to better achieve desired behaviors. However, tuning these models has become increasingly resource-intensive, or impossible when model weights are private. We introduce proxy-tuning, a lightweight decoding-time algorithm that operates on top of black-box LMs to achieve the result of directly tuning the model, but by accessing only its prediction over the output vocabulary. Our method instead tunes a smaller LM, then applies the difference between the predictions of the small tuned and untuned LMs to shift the original predictions of the base model in the direction of tuning, while retaining the benefits of larger scale pretraining. In experiments, when we apply proxy-tuning to Llama2-70B using proxies of only 7B size, we can close 88% of the gap between Llama2-70B and its truly-tuned chat version, when evaluated across knowledge, reasoning, and safety benchmarks. Interestingly, when tested on TruthfulQA, proxy-tuned models are actually more truthful than directly tuned models, possibly because decoding-time guidance better retains the model's factual knowledge. We then demonstrate the generality of proxy-tuning by applying it for domain adaptation on code, and task-specific finetuning on question-answering and math problems. Our work demonstrates the promise of using small tuned LMs to efficiently customize large, potentially proprietary LMs through decoding-time guidance.

This enables immediate control over vehicle functions like brakes, acceleration, and steering. It offers advantages such as contributing to traffic safety by delivering precise information directly to drivers. However, the dynamic nature of VANETs, marked by constantly changing network topologies, varying vehicle speeds, and differences in the density of V2X communications, introduces new challenges and vulnerabilities that must be addressed [1]. These vulnerabilities can be exploited to launch various types of attacks, which could result in various issues such as accidents and traffic congestion. Thus, ensuring the security of VANETs is of great significance due to the potential risks to human lives, property, and economic activities. This underscores the need to prioritize the development of robust information system security tools and mechanisms capable of not only detecting but also effectively mitigating these attacks. Taking proactive measures is essential to ensure the integrity and safety of VANETs in the face of the evolving cybersecurity threats.

Exploration of the lunar south pole with a solar-powered rover is challenging due to the highly dynamic solar illumination conditions and the presence of permanently shadowed regions (PSRs). In turn, careful planning in space and time is essential. Mission-level path planning is a global, spatiotemporal paradigm that addresses this challenge, taking into account rover resources and mission requirements. However, existing approaches do not proactively account for random disturbances, such as recurring faults, that may temporarily delay rover traverse progress. In this paper, we formulate a chance-constrained mission-level planning problem for the exploration of PSRs by a solar-powered rover affected by random faults. The objective is to find a policy that visits as many waypoints of scientific interest as possible while respecting an upper bound on the probability of mission failure. Our approach assumes that faults occur randomly, but at a known, constant average rate. Each fault is resolved within a fixed time, simulating the recovery period of an autonomous system or the time required for a team of human operators to intervene. Unlike solutions based upon dynamic programming alone, our method breaks the chance-constrained optimization problem into smaller offline and online subtasks to make the problem computationally tractable. Specifically, our solution combines existing mission-level path planning techniques with a stochastic reachability analysis component. We find mission plans that remain within reach of safety throughout large state spaces. To empirically validate our algorithm, we simulate mission scenarios using orbital terrain and illumination maps of Cabeus Crater. Results from simulations of multi-day, long-range drives in the LCROSS impact region are also presented.

This paper provides an overview of the morphology and syntax of the Tamil language, focusing on its contemporary usage. The paper also highlights the complexity and richness of Tamil in terms of its morphological and syntactic features, which will be useful for linguists analysing the language and conducting comparative studies. In addition, the paper will be useful for those developing computational resources for the Tamil language. It is proven as a rule-based morphological analyser cum generator and a computational grammar for Tamil have already been developed based on this paper. To enhance accessibility for a broader audience, the analysis is conducted without relying on any specific grammatical formalism.

Moreover, PRs have other advantages over their serial counterparts, such as lower weight, higher working speed with high precision, and lower power consumption [1,2]. These advantages, mainly due to the closed kinematic chain architecture, are key aspects that have increased the interest in studying their use in the academic, industrial, and robotics service fields over the last three decades. However, the PR architecture reduces not only the size of the robot workspace but also its kinematic performance, owing to the possible presence of singularities within the workspace. Initially, Gosselin and Angeles [3] studied the singularities of a PR using Jacobian matrices obtained from constraint equations, and classified them into i) inverse kinematic or Type I singularity, where the robot loses at least one degree of freedom (DOF), and ii) Forward Kinematic or Type II singularity, where the PR gains at least one DOF. In particular, Type II singularities could be critical because the mobile platform at the singularity is unable to bear the external forces despite having all the actuators locked (losing control of the PR motion).

Although parallel manipulators (PMs) started with the introduction of architectures with 6 Degrees of Freedom (DoF), a vast number of applications require less than 6 DoF. Consequently, scholars have proposed architectures with 3 DoF and 4 DoF, but relatively few 4 DoF PMs have become prototypes, especially of the two rotation (2R) and two translation (2T) motion types. In this paper, we explain the mechatronics design, prototype and control architecture design of a 4 DoF PM with 2R2T motions. We chose to design a 4 DoF manipulator based on the motion needed to complete the tasks of lower limb rehabilitation. To the author's best knowledge, PMs between 3 and 6 DoF for rehabilitation of lower limbs have not been proposed to date. The developed architecture enhances the three minimum DoF required by adding a 4 DoF which allows combinations of normal or tangential efforts in the joints, or torque acting on the knee. We put forward the inverse and forward displacement equations, describe the prototype, perform the experimental setup, and develop the hardware and control architecture. The tracking accuracy experiments from the proposed controller show that the manipulator can accomplish the required application.

This paper aims to develop an approach for the reconfiguration of a parallel kinematic manipulator (PKM) with four degrees of freedom (DoF) designed to tackle tasks of diagnosis and rehabilitation in an injured knee. The original layout of the 4-DoF manipulator presents Type-II singular configurations within its workspace. Thus, we proposed to reconfigure the manipulator to avoid such singularities (owing to the Forward Jacobian of the PKM) during typical rehabilitation trajectories. We achieve the reconfiguration of the PKM through a minimization problem where the design variables correspond to the anchoring points of the robot limbs on fixed and mobile platforms. The objective function relies on the minimization of the forces exerted by the actuators for a specific trajectory. The minimization problem considers constraint equations to avoid Type-II singularities, which guarantee the feasibility of the active generalized coordinates for a particular path. To evaluate the proposed conceptual strategy, we build a prototype where reconfiguration occurs by moving the position of the anchoring points to holes bored in the fixed and mobile platforms. Simulations and experiments of several study cases enable testing the strategy performance. The results show that the reconfiguration strategy allows obtaining trajectories having minimum actuation forces without Type-II singularities.

Knowledge graphs (KGs) often contain various errors. Previous works on detecting errors in KGs mainly rely on triplet embedding from graph structure. We conduct an empirical study and find that these works struggle to discriminate noise from semantically-similar correct triplets. In this paper, we propose a KG error detection model CCA to integrate both textual and graph structural information from triplet reconstruction for better distinguishing semantics. We design interactive contrastive learning to capture the differences between textual and structural patterns. Furthermore, we construct realistic datasets with semantically-similar noise and adversarial noise. Experimental results demonstrate that CCA outperforms state-of-the-art baselines, especially in detecting semantically-similar noise and adversarial noise.

Multimodal large language models (MLLMs) have shown remarkable capabilities across a broad range of tasks but their knowledge and abilities in the geographic and geospatial domains are yet to be explored, despite potential wide-ranging benefits to navigation, environmental research, urban development, and disaster response. We conduct a series of experiments exploring various vision capabilities of MLLMs within these domains, particularly focusing on the frontier model GPT-4V, and benchmark its performance against open-source counterparts. Our methodology involves challenging these models with a small-scale geographic benchmark consisting of a suite of visual tasks, testing their abilities across a spectrum of complexity. The analysis uncovers not only where such models excel, including instances where they outperform humans, but also where they falter, providing a balanced view of their capabilities in the geographic domain. To enable the comparison and evaluation of future models, our benchmark will be publicly released.

Artificial intelligence will affect 40% of jobs around the world and it is "crucial" that countries build social safety nets to mitigate the impact on vulnerable workers, according to the head of the International Monetary Fund. AI, the term for computer systems that can perform tasks usually associated with human levels of intelligence, is poised to profoundly change the global economy with advanced economies at greater risk of disruption. Analysis by the IMF, the international lender of last resort, says about 60% of jobs in advanced economies such as the US and UK are exposed to AI and half of these jobs may be negatively affected. But the technology will also help to enhance some humans' productivity as AI improves their performance, it said. According to the IMF, the safest highly exposed jobs are those with a "high complementarity" to AI, meaning the technology will assist their work rather than displace it entirely.