Transformer-based large language models (LLMs) exhibit impressive performance in generative tasks but introduce significant challenges in real-world serving due to inefficient use of the expensive, computation-optimized accelerators. This mismatch arises from the autoregressive nature of LLMs, where the generation phase comprises operators with varying resource demands. Specifically, the attention operator is memory-intensive, exhibiting a memory access pattern that clashes with the strengths of modern accelerators, especially as context length increases. To enhance the efficiency and cost-effectiveness of LLM serving, we introduce the concept of attention offloading. This approach leverages a collection of cheap, memory-optimized devices for the attention operator while still utilizing high-end accelerators for other parts of the model. This heterogeneous setup ensures that each component is tailored to its specific workload, maximizing overall performance and cost efficiency. Our comprehensive analysis and experiments confirm the viability of splitting the attention computation over multiple devices. Also, the communication bandwidth required between heterogeneous devices proves to be manageable with prevalent networking technologies. To further validate our theory, we develop Lamina, an LLM inference system that incorporates attention offloading. Experimental results indicate that Lamina can provide 1.48x-12.1x higher estimated throughput per dollar than homogeneous solutions.

Large 2D vision-language models (2D-LLMs) have gained significant attention by bridging Large Language Models (LLMs) with images using a simple projector. Inspired by their success, large 3D point cloud-language models (3D-LLMs) also integrate point clouds into LLMs. However, directly aligning point clouds with LLM requires expensive training costs, typically in hundreds of GPU-hours on A100, which hinders the development of 3D-LLMs. In this paper, we introduce MiniGPT-3D, an efficient and powerful 3D-LLM that achieves multiple SOTA results while training for only 27 hours on one RTX 3090. Specifically, we propose to align 3D point clouds with LLMs using 2D priors from 2D-LLMs, which can leverage the similarity between 2D and 3D visual information. We introduce a novel four-stage training strategy for modality alignment in a cascaded way, and a mixture of query experts module to adaptively aggregate features with high efficiency. Moreover, we utilize parameter-efficient fine-tuning methods LoRA and Norm fine-tuning, resulting in only 47.8M learnable parameters, which is up to 260x fewer than existing methods. Extensive experiments show that MiniGPT-3D achieves SOTA on 3D object classification and captioning tasks, with significantly cheaper training costs. Notably, MiniGPT-3D gains an 8.12 increase on GPT-4 evaluation score for the challenging object captioning task compared to ShapeLLM-13B, while the latter costs 160 total GPU-hours on 8 A800. We are the first to explore the efficient 3D-LLM, offering new insights to the community. Code and weights are available at https://github.com/TangYuan96/MiniGPT-3D.

Mobility analysis is a crucial element in the research area of transportation systems. Forecasting traffic information offers a viable solution to address the conflict between increasing transportation demands and the limitations of transportation infrastructure. Predicting human travel is significant in aiding various transportation and urban management tasks, such as taxi dispatch and urban planning. Machine learning and deep learning methods are favored for their flexibility and accuracy. Nowadays, with the advent of large language models (LLMs), many researchers have combined these models with previous techniques or applied LLMs to directly predict future traffic information and human travel behaviors. However, there is a lack of comprehensive studies on how LLMs can contribute to this field. This survey explores existing approaches using LLMs for mobility forecasting problems. We provide a literature review concerning the forecasting applications within transportation systems, elucidating how researchers utilize LLMs, showcasing recent state-of-the-art advancements, and identifying the challenges that must be overcome to fully leverage LLMs in this domain.

A world model creates a surrogate world to train a controller and predict safety violations by learning the internal dynamic model of systems. However, the existing world models rely solely on statistical learning of how observations change in response to actions, lacking precise quantification of how accurate the surrogate dynamics are, which poses a significant challenge in safety-critical systems. To address this challenge, we propose foundation world models that embed observations into meaningful and causally latent representations. This enables the surrogate dynamics to directly predict causal future states by leveraging a training-free large language model. In two common benchmarks, this novel model outperforms standard world models in the safety prediction task and has a performance comparable to supervised learning despite not using any data. We evaluate its performance with a more specialized and system-relevant metric by comparing estimated states instead of aggregating observation-wide error.

Unmanned Aerial Vehicles (UAVs) have emerged as a transformative technology across diverse sectors, offering adaptable solutions to complex challenges in both military and civilian domains. Their expanding capabilities present a platform for further advancement by integrating cutting-edge computational tools like Artificial Intelligence (AI) and Machine Learning (ML) algorithms. These advancements have significantly impacted various facets of human life, fostering an era of unparalleled efficiency and convenience. Large Language Models (LLMs), a key component of AI, exhibit remarkable learning and adaptation capabilities within deployed environments, demonstrating an evolving form of intelligence with the potential to approach human-level proficiency. This work explores the significant potential of integrating UAVs and LLMs to propel the development of autonomous systems. We comprehensively review LLM architectures, evaluating their suitability for UAV integration. Additionally, we summarize the state-of-the-art LLM-based UAV architectures and identify novel opportunities for LLM embedding within UAV frameworks. Notably, we focus on leveraging LLMs to refine data analysis and decision-making processes, specifically for enhanced spectral sensing and sharing in UAV applications. Furthermore, we investigate how LLM integration expands the scope of existing UAV applications, enabling autonomous data processing, improved decision-making, and faster response times in emergency scenarios like disaster response and network restoration. Finally, we highlight crucial areas for future research that are critical for facilitating the effective integration of LLMs and UAVs.

The training of Transformer models has revolutionized natural language processing and computer vision, but it remains a resource-intensive and time-consuming process. This paper investigates the applicability of the early-bird ticket hypothesis to optimize the training efficiency of Transformer models. We propose a methodology that combines iterative pruning, masked distance calculation, and selective retraining to identify early-bird tickets in various Transformer architectures, including ViT, Swin-T, GPT-2, and RoBERTa. Our experimental results demonstrate that early-bird tickets can be consistently found within the first few epochs of training or fine-tuning, enabling significant resource optimization without compromising performance. The pruned models obtained from early-bird tickets achieve comparable or even superior accuracy to their unpruned counterparts while substantially reducing memory usage. Furthermore, our comparative analysis highlights the generalizability of the early-bird ticket phenomenon across different Transformer models and tasks. This research contributes to the development of efficient training strategies for Transformer models, making them more accessible and resource-friendly. By leveraging early-bird tickets, practitioners can accelerate the progress of natural language processing and computer vision applications while reducing the computational burden associated with training Transformer models.

Utilizing large language models to generate codes has shown promising meaning in software development revolution. Despite the intelligence shown by the general large language models, their specificity in code generation can still be improved due to the syntactic gap and mismatched vocabulary existing among natural language and different programming languages. In addition, programming languages are inherently logical and complex, making them hard to be correctly generated. Existing methods rely on multiple prompts to the large language model to explore better solutions, which is expensive. In this paper, we propose Syntax Graph Retrieval Augmented Code Generation (CodeGRAG) to enhance the performance of LLMs in single-round code generation tasks. CodeGRAG extracts and summarizes the control flow and data flow of code blocks to fill the gap between programming languages and natural language. The extracted external structural knowledge models the inherent flows of code blocks, which can facilitate LLMs for better understanding of code syntax and serve as a bridge among different programming languages. CodeGRAG significantly improves the code generation ability of LLMs and can even offer performance gain for cross-lingual code generation, e.g., C++ for Python.

Diagnosing language disorders associated with autism is a complex and nuanced challenge, often hindered by the subjective nature and variability of traditional assessment methods. Traditional diagnostic methods not only require intensive human effort but also often result in delayed interventions due to their lack of speed and specificity. In this study, we explored the application of ChatGPT, a state of the art large language model, to overcome these obstacles by enhancing diagnostic accuracy and profiling specific linguistic features indicative of autism. Leveraging ChatGPT advanced natural language processing capabilities, this research aims to streamline and refine the diagnostic process. Specifically, we compared ChatGPT's performance with that of conventional supervised learning models, including BERT, a model acclaimed for its effectiveness in various natural language processing tasks. We showed that ChatGPT substantially outperformed these models, achieving over 13% improvement in both accuracy and F1 score in a zero shot learning configuration. This marked enhancement highlights the model potential as a superior tool for neurological diagnostics. Additionally, we identified ten distinct features of autism associated language disorders that vary significantly across different experimental scenarios. These features, which included echolalia, pronoun reversal, and atypical language usage, were crucial for accurately diagnosing ASD and customizing treatment plans. Together, our findings advocate for adopting sophisticated AI tools like ChatGPT in clinical settings to assess and diagnose developmental disorders. Our approach not only promises greater diagnostic precision but also aligns with the goals of personalized medicine, potentially transforming the evaluation landscape for autism and similar neurological conditions.

We seek to learn a generalizable goal-conditioned policy that enables zero-shot robot manipulation: interacting with unseen objects in novel scenes without test-time adaptation. While typical approaches rely on a large amount of demonstration data for such generalization, we propose an approach that leverages web videos to predict plausible interaction plans and learns a task-agnostic transformation to obtain robot actions in the real world. Our framework,Track2Act predicts tracks of how points in an image should move in future time-steps based on a goal, and can be trained with diverse videos on the web including those of humans and robots manipulating everyday objects. We use these 2D track predictions to infer a sequence of rigid transforms of the object to be manipulated, and obtain robot end-effector poses that can be executed in an open-loop manner. We then refine this open-loop plan by predicting residual actions through a closed loop policy trained with a few embodiment-specific demonstrations. We show that this approach of combining scalably learned track prediction with a residual policy requiring minimal in-domain robot-specific data enables zero-shot robot manipulation, and present a wide array of real-world robot manipulation results across unseen tasks, objects, and scenes. https://homangab.github.io/track2act/

It turns out the lay of today's AI landscape can be traced back to -- what do you know -- fear, jealousy and intense capitalist ambition. Emails revealed in the Department of Justice's antitrust case against Google, first reported by Business Insider, show Microsoft executives expressing alarm and envy over Google's AI lead. That spurred an urgency that led to the Windows maker's initial billion-dollar investment in its now-indispensable partner, OpenAI. In a heavily redacted 2019 email thread titled "Thoughts on OpenAI," Microsoft CEO Satya Nadella forwards a lengthy message from CTO Kevin Scott to CFO Amy Hood. "Very good email that explains, why I want us to do this ... and also why we will then ensure our infra folks execute," Nadella wrote.