With the rapid advancement of artificial intelligence technologies such as ChatGPT, AI agents and video generation,contemporary mobile systems have begun integrating these AI capabilities on local devices to enhance privacy and reduce response latency. To meet the computational demands of AI tasks, current mobile SoCs are equipped with diverse AI accelerators, including GPUs and Neural Processing Units (NPUs). However, there has not been a comprehensive characterization of these heterogeneous processors, and existing designs typically only leverage a single AI accelerator for LLM inference, leading to suboptimal use of computational resources and memory bandwidth. In this paper, we first summarize key performance characteristics of mobile SoC, including heterogeneous processors, unified memory, synchronization, etc. Drawing on these observations, we propose different tensor partition strategies to fulfill the distinct requirements of the prefill and decoding phases. We further design a fast synchronization mechanism that leverages the unified memory address provided by mobile SoCs. By employing these techniques, we present HeteroLLM, the fastest LLM inference engine in mobile devices which supports both layer-level and tensor-level heterogeneous execution. Evaluation results show that HeteroLLM achieves 9.99 and 4.36 performance improvement over other mobile-side LLM inference engines: MLC and MNN.
The rapid growth of generative AI and its integration into everyday workflows have significantly increased the demand for large language model (LLM) inference services. While proprietary models remain popular, recent advancements in open-source LLMs have positioned them as strong contenders. However, deploying these models is often constrained by the high costs and limited availability of GPU resources. In response, this paper presents the design of a decentralized offline serving system for LLM inference. Utilizing idle GPU resources, our proposed system, DeServe, decentralizes access to LLMs at a lower cost. DeServe specifically addresses key challenges in optimizing serving throughput in high-latency network environments. Experiments demonstrate that DeServe achieves a 6.7x-12.6x improvement in throughput over existing serving system baselines in such conditions.
In this paper, we present a system that uses a Large Language Model (LLM) to perform grammar and spelling correction as a component of Quality Assurance (QA) for texts generated by NLG systems, which is important for text production in real-world scenarios. Evaluating the results of the system on work-in-progress sports news texts in three languages, we show that it is able to deliver acceptable corrections.
Endangered languages, such as Navajo - the most widely spoken Native American language - are significantly underrepresented in contemporary language technologies, exacerbating the challenges of their preservation and revitalization. This study evaluates Google's large language model (LLM)-based language identification system, which consistently misidentifies Navajo, exposing inherent limitations when applied to low-resource Native American languages. To address this, we introduce a random forest classifier trained on Navajo and eight frequently confused languages. Despite its simplicity, the classifier achieves near-perfect accuracy (97-100%), significantly outperforming Google's LLM-based system. Additionally, the model demonstrates robustness across other Athabaskan languages - a family of Native American languages spoken primarily in Alaska, the Pacific Northwest, and parts of the Southwestern United States - suggesting its potential for broader application. Our findings underscore the pressing need for NLP systems that prioritize linguistic diversity and adaptability over centralized, one-size-fits-all solutions, especially in supporting underrepresented languages in a multicultural world. This work directly contributes to ongoing efforts to address cultural biases in language models and advocates for the development of culturally localized NLP tools that serve diverse linguistic communities.
This study introduces an approach to Estonian text simplification using two model architectures: a neural machine translation model and a fine-tuned large language model (LLaMA). Given the limited resources for Estonian, we developed a new dataset, the Estonian Simplification Dataset, combining translated data and GPT-4.0-generated simplifications. We benchmarked OpenNMT, a neural machine translation model that frames text simplification as a translation task, and fine-tuned the LLaMA model on our dataset to tailor it specifically for Estonian simplification. Manual evaluations on the test set show that the LLaMA model consistently outperforms OpenNMT in readability, grammaticality, and meaning preservation. These findings underscore the potential of large language models for low-resource languages and provide a basis for further research in Estonian text simplification.
In human-centric scenes, the ability to simultaneously understand visual and auditory information is crucial. While recent omni models can process multiple modalities, they generally lack effectiveness in human-centric scenes due to the absence of large-scale, specialized datasets and non-targeted architectures. In this work, we developed HumanOmni, the industry's first human-centric Omni-multimodal large language model. We constructed a dataset containing over 2.4 million human-centric video clips with detailed captions and more than 14 million instructions, facilitating the understanding of diverse human-centric scenes. HumanOmni includes three specialized branches for understanding different types of scenes. It adaptively fuses features from these branches based on user instructions, significantly enhancing visual understanding in scenes centered around individuals. Moreover, HumanOmni integrates audio features to ensure a comprehensive understanding of environments and individuals. Our experiments validate HumanOmni's advanced capabilities in handling human-centric scenes across a variety of tasks, including emotion recognition, facial expression description, and action understanding. Our model will be open-sourced to facilitate further development and collaboration within both academia and industry.
The analysis of extended video content poses unique challenges in artificial intelligence, particularly when dealing with the complexity of tracking and understanding visual elements across time. Current methodologies that process video frames sequentially struggle to maintain coherent tracking of objects, especially when these objects temporarily vanish and later reappear in the footage. A critical limitation of these approaches is their inability to effectively identify crucial moments in the video, largely due to their limited grasp of temporal relationships. To overcome these obstacles, we present GraphVideoAgent, a cutting-edge system that leverages the power of graph-based object tracking in conjunction with large language model capabilities. At its core, our framework employs a dynamic graph structure that maps and monitors the evolving relationships between visual entities throughout the video sequence. This innovative approach enables more nuanced understanding of how objects interact and transform over time, facilitating improved frame selection through comprehensive contextual awareness. Our approach demonstrates remarkable effectiveness when tested against industry benchmarks. In evaluations on the EgoSchema dataset, GraphVideoAgent achieved a 2.2 improvement over existing methods while requiring analysis of only 8.2 frames on average. Similarly, testing on the NExT-QA benchmark yielded a 2.0 performance increase with an average frame requirement of 8.1. These results underscore the efficiency of our graph-guided methodology in enhancing both accuracy and computational performance in long-form video understanding tasks.
In view of the gap in the current large language model in sharing memory across dialogues, this research proposes a wormhole memory module (WMM) to realize memory as a Rubik's cube that can be arbitrarily retrieved between different dialogues. Through simulation experiments, the researcher built an experimental framework based on the Python environment and used setting memory barriers to simulate the current situation where memories between LLMs dialogues are difficult to share. The CoQA development data set was imported into the experiment, and the feasibility of its cross-dialogue memory retrieval function was verified for WMM's nonlinear indexing and dynamic retrieval, and a comparative analysis was conducted with the capabilities of Titans and MemGPT memory modules. Experimental results show that WMM demonstrated the ability to retrieve memory across dialogues and the stability of quantitative indicators in eight experiments. It contributes new technical approaches to the optimization of memory management of LLMs and provides experience for the practical application in the future.
Interactive Text-to-Image Retrieval (I-TIR) has emerged as a transformative user-interactive tool for applications in domains such as e-commerce and education. Yet, current methodologies predominantly depend on finetuned Multimodal Large Language Models (MLLMs), which face two critical limitations: (1) Finetuning imposes prohibitive computational overhead and long-term maintenance costs. (2) Finetuning narrows the pretrained knowledge distribution of MLLMs, reducing their adaptability to novel scenarios. These issues are exacerbated by the inherently dynamic nature of real-world I-TIR systems, where queries and image databases evolve in complexity and diversity, often deviating from static training distributions. To overcome these constraints, we propose Diffusion Augmented Retrieval (DAR), a paradigm-shifting framework that bypasses MLLM finetuning entirely. DAR synergizes Large Language Model (LLM)-guided query refinement with Diffusion Model (DM)-based visual synthesis to create contextually enriched intermediate representations. This dual-modality approach deciphers nuanced user intent more holistically, enabling precise alignment between textual queries and visually relevant images. Rigorous evaluations across four benchmarks reveal DAR's dual strengths: (1) Matches state-of-the-art finetuned I-TIR models on straightforward queries without task-specific training. (2) Scalable Generalization: Surpasses finetuned baselines by 7.61% in Hits@10 (top-10 accuracy) under multi-turn conversational complexity, demonstrating robustness to intricate, distributionally shifted interactions. By eliminating finetuning dependencies and leveraging generative-augmented representations, DAR establishes a new trajectory for efficient, adaptive, and scalable cross-modal retrieval systems.
Writing comprehensive and accurate descriptions of technical drawings in patent documents is crucial to effective knowledge sharing and enabling the replication and protection of intellectual property. However, automation of this task has been largely overlooked by the research community. To this end, we introduce PatentDesc-355K, a novel large-scale dataset containing ~355K patent figures along with their brief and detailed textual descriptions extracted from more than 60K US patent documents. In addition, we propose PatentLMM - a novel multimodal large language model specifically tailored to generate high-quality descriptions of patent figures. Our proposed PatentLMM comprises two key components: (i) PatentMME, a specialized multimodal vision encoder that captures the unique structural elements of patent figures, and (ii) PatentLLaMA, a domain-adapted version of LLaMA fine-tuned on a large collection of patents. Extensive experiments demonstrate that training a vision encoder specifically designed for patent figures significantly boosts the performance, generating coherent descriptions compared to fine-tuning similar-sized off-the-shelf multimodal models. PatentDesc-355K and PatentLMM pave the way for automating the understanding of patent figures, enabling efficient knowledge sharing and faster drafting of patent documents. We make the code and data publicly available.
Power distribution networks are evolving due to the integration of DERs and increased customer participation. To maintain optimal operation, minimize losses, and meet varying load demands, frequent network reconfiguration is necessary. Traditionally, the reconfiguration task relies on optimization software and expert operators, but as systems grow more complex, faster and more adaptive solutions are required without expert intervention. Data-driven reconfiguration is gaining traction for its accuracy, speed, and robustness against incomplete network data. LLMs, with their ability to capture complex patterns, offer a promising approach for efficient and responsive network reconfiguration in evolving complex power networks. In this work, we introduce LLM4DistReconfig, a deep learning-based approach utilizing a fine-tuned LLM to solve the distribution network reconfiguration problem. By carefully crafting prompts and designing a custom loss function, we train the LLM with inputs representing network parameters such as buses, available lines, open lines, node voltages, and system loss. The model then predicts optimal reconfigurations by outputting updated network configurations that minimize system loss while meeting operational constraints. Our approach significantly reduces inference time compared to classical algorithms, allowing for near real-time optimal reconfiguration after training. Experimental results show that our method generates optimal configurations minimizing system loss for five individual and a combined test dataset. It also produces minimal invalid edges, no cycles, or subgraphs across all datasets, fulfilling domain-specific needs. Additionally, the generated responses contain less than 5% improper outputs on seen networks and satisfactory results on unseen networks, demonstrating its effectiveness and reliability for the reconfiguration task.
Purpose: This study aims to use a large language model (LLM) to automate the generation of summaries from the CT simulation orders and evaluate its performance. Materials and Methods: A total of 607 CT simulation orders for patients were collected from the Aria database at our institution. A locally hosted Llama 3.1 405B model, accessed via the Application Programming Interface (API) service, was used to extract keywords from the CT simulation orders and generate summaries. The downloaded CT simulation orders were categorized into seven groups based on treatment modalities and disease sites. For each group, a customized instruction prompt was developed collaboratively with therapists to guide the Llama 3.1 405B model in generating summaries. The ground truth for the corresponding summaries was manually derived by carefully reviewing each CT simulation order and subsequently verified by therapists. The accuracy of the LLM-generated summaries was evaluated by therapists using the verified ground truth as a reference. Results: About 98% of the LLM-generated summaries aligned with the manually generated ground truth in terms of accuracy. Our evaluations showed an improved consistency in format and enhanced readability of the LLM-generated summaries compared to the corresponding therapists-generated summaries. This automated approach demonstrated a consistent performance across all groups, regardless of modality or disease site. Conclusions: This study demonstrated the high precision and consistency of the Llama 3.1 405B model in extracting keywords and summarizing CT simulation orders, suggesting that LLMs have great potential to help with this task, reduce the workload of therapists and improve workflow efficiency.
Protecting online privacy requires users to engage with and comprehend website privacy policies, but many policies are difficult and tedious to read. We present PRISMe (Privacy Risk Information Scanner for Me), a novel Large Language Model (LLM)-driven privacy policy assessment tool, which helps users to understand the essence of a lengthy, complex privacy policy while browsing. The tool, a browser extension, integrates a dashboard and an LLM chat. One major contribution is the first rigorous evaluation of such a tool. In a mixed-methods user study (N=22), we evaluate PRISMe's efficiency, usability, understandability of the provided information, and impacts on awareness. While our tool improves privacy awareness by providing a comprehensible quick overview and a quality chat for in-depth discussion, users note issues with consistency and building trust in the tool. From our insights, we derive important design implications to guide future policy analysis tools.
Automated grading has become an essential tool in education technology due to its ability to efficiently assess large volumes of student work, provide consistent and unbiased evaluations, and deliver immediate feedback to enhance learning. However, current systems face significant limitations, including the need for large datasets in few-shot learning methods, a lack of personalized and actionable feedback, and an overemphasis on benchmark performance rather than student experience. To address these challenges, we propose a Zero-Shot Large Language Model (LLM)-Based Automated Assignment Grading (AAG) system. This framework leverages prompt engineering to evaluate both computational and explanatory student responses without requiring additional training or fine-tuning. The AAG system delivers tailored feedback that highlights individual strengths and areas for improvement, thereby enhancing student learning outcomes. Our study demonstrates the system's effectiveness through comprehensive evaluations, including survey responses from higher education students that indicate significant improvements in motivation, understanding, and preparedness compared to traditional grading methods. The results validate the AAG system's potential to transform educational assessment by prioritizing learning experiences and providing scalable, high-quality feedback.
We introduce Zep, a novel memory layer service for AI agents that outperforms the current state-of-the-art system, MemGPT, in the Deep Memory Retrieval (DMR) benchmark. Additionally, Zep excels in more comprehensive and challenging evaluations than DMR that better reflect real-world enterprise use cases. While existing retrieval-augmented generation (RAG) frameworks for large language model (LLM)-based agents are limited to static document retrieval, enterprise applications demand dynamic knowledge integration from diverse sources including ongoing conversations and business data. Zep addresses this fundamental limitation through its core component Graphiti -- a temporally-aware knowledge graph engine that dynamically synthesizes both unstructured conversational data and structured business data while maintaining historical relationships. In the DMR benchmark, which the MemGPT team established as their primary evaluation metric, Zep demonstrates superior performance (94.8% vs 93.4%). Beyond DMR, Zep's capabilities are further validated through the more challenging LongMemEval benchmark, which better reflects enterprise use cases through complex temporal reasoning tasks. In this evaluation, Zep achieves substantial results with accuracy improvements of up to 18.5% while simultaneously reducing response latency by 90% compared to baseline implementations. These results are particularly pronounced in enterprise-critical tasks such as cross-session information synthesis and long-term context maintenance, demonstrating Zep's effectiveness for deployment in real-world applications.
We introduce DRESS, a novel approach for generating stylized large language model (LLM) responses through representation editing. Existing methods like prompting and fine-tuning are either insufficient for complex style adaptation or computationally expensive, particularly in tasks like NPC creation or character role-playing. Our approach leverages the over-parameterized nature of LLMs to disentangle a style-relevant subspace within the model's representation space to conduct representation editing, ensuring a minimal impact on the original semantics. By applying adaptive editing strengths, we dynamically adjust the steering vectors in the style subspace to maintain both stylistic fidelity and semantic integrity. We develop two stylized QA benchmark datasets to validate the effectiveness of DRESS, and the results demonstrate significant improvements compared to baseline methods such as prompting and ITI. In short, DRESS is a lightweight, train-free solution for enhancing LLMs with flexible and effective style control, making it particularly useful for developing stylized conversational agents. Codes and benchmark datasets are available at https://github.com/ArthurLeoM/DRESS-LLM.
Large language model (LLM) inference workload dominates a wide variety of modern AI applications, ranging from multi-turn conversation to document analysis. Balancing fairness and efficiency is critical for managing diverse client workloads with varying prefix patterns. Unfortunately, existing fair scheduling algorithms for LLM serving, such as Virtual Token Counter (VTC), fail to take prefix locality into consideration and thus suffer from poor performance. On the other hand, locality-aware scheduling algorithms in existing LLM serving frameworks tend to maximize the prefix cache hit rate without considering fair sharing among clients. This paper introduces the first locality-aware fair scheduling algorithm, Deficit Longest Prefix Match (DLPM), which can maintain a high degree of prefix locality with a fairness guarantee. We also introduce a novel algorithm, Double Deficit LPM (D$^2$LPM), extending DLPM for the distributed setup that can find a balance point among fairness, locality, and load-balancing. Our extensive evaluation demonstrates the superior performance of DLPM and D$^2$LPM in ensuring fairness while maintaining high throughput (up to 2.87$\times$ higher than VTC) and low per-client (up to 7.18$\times$ lower than state-of-the-art distributed LLM serving system) latency.
Graph Retrieval Augmented Generation (GRAG) is a novel paradigm that takes the naive RAG system a step further by integrating graph information, such as knowledge graph (KGs), into large-scale language models (LLMs) to mitigate hallucination. However, existing GRAG still encounter limitations: 1) simple paradigms usually fail with the complex problems due to the narrow and shallow correlations capture from KGs 2) methods of strong coupling with KGs tend to be high computation cost and time consuming if the graph is dense. In this paper, we propose the Fast Think-on-Graph (FastToG), an innovative paradigm for enabling LLMs to think ``community by community" within KGs. To do this, FastToG employs community detection for deeper correlation capture and two stages community pruning - coarse and fine pruning for faster retrieval. Furthermore, we also develop two Community-to-Text methods to convert the graph structure of communities into textual form for better understanding by LLMs. Experimental results demonstrate the effectiveness of FastToG, showcasing higher accuracy, faster reasoning, and better explainability compared to the previous works.
The effectiveness of search systems is evaluated using relevance labels that indicate the usefulness of documents for specific queries and users. While obtaining these relevance labels from real users is ideal, scaling such data collection is challenging. Consequently, third-party annotators are employed, but their inconsistent accuracy demands costly auditing, training, and monitoring. We propose an LLM-based modular classification pipeline that divides the relevance assessment task into multiple stages, each utilising different prompts and models of varying sizes and capabilities. Applied to TREC Deep Learning (TREC-DL), one of our approaches showed an 18.4% Krippendorff's $\alpha$ accuracy increase over OpenAI's GPT-4o mini while maintaining a cost of about 0.2 USD per million input tokens, offering a more efficient and scalable solution for relevance assessment. This approach beats the baseline performance of GPT-4o (5 USD). With a pipeline approach, even the accuracy of the GPT-4o flagship model, measured in $\alpha$, could be improved by 9.7%.
Benchmarks are important tools for tracking the rapid advancements in large language model (LLM) capabilities. However, benchmarks are not keeping pace in difficulty: LLMs now achieve over 90\% accuracy on popular benchmarks like MMLU, limiting informed measurement of state-of-the-art LLM capabilities. In response, we introduce Humanity's Last Exam (HLE), a multi-modal benchmark at the frontier of human knowledge, designed to be the final closed-ended academic benchmark of its kind with broad subject coverage. HLE consists of 2,500 questions across dozens of subjects, including mathematics, humanities, and the natural sciences. HLE is developed globally by subject-matter experts and consists of multiple-choice and short-answer questions suitable for automated grading. Each question has a known solution that is unambiguous and easily verifiable, but cannot be quickly answered via internet retrieval. State-of-the-art LLMs demonstrate low accuracy and calibration on HLE, highlighting a significant gap between current LLM capabilities and the expert human frontier on closed-ended academic questions. To inform research and policymaking upon a clear understanding of model capabilities, we publicly release HLE at https://lastexam.ai.