arXiv subjects

Jun Zhao

Publications and source records attributed to Jun Zhao.

At least 253 records · Page 14Linked to original sources

arXiv2024

Play to Earn in the Metaverse with Mobile Edge Computing over Wireless Networks: A Deep Reinforcement Learning Approach

The Metaverse play-to-earn games have been gaining popularity as they enable players to earn in-game tokens which can be translated to real-world profits. With the advancements in augmented reality (AR) technologies, users can play AR games in the Metaverse. However, these high-resolution games are compute-intensive, and in-game graphical scenes need to be offloaded from mobile devices to an edge server for computation. In this work, we consider an optimization problem where the Metaverse Service Provider (MSP)'s objective is to reduce downlink transmission latency of in-game graphics, the latency of uplink data transmission, and the worst-case (greatest) battery charge expenditure of user equipments (UEs), while maximizing the worst-case (lowest) UE resolution-influenced in-game earning potential through optimizing the downlink UE-Metaverse Base Station (UE-MBS) assignment and the uplink transmission power selection. The downlink and uplink transmissions are then executed asynchronously. We propose a multi-agent, loss-sharing (MALS) reinforcement learning model to tackle the asynchronous and asymmetric problem. We then compare the MALS model with other baseline models and show its superiority over other methods. Finally, we conduct multi-variable optimization weighting analyses and show the viability of using our proposed MALS algorithm to tackle joint optimization problems.

cs.NI↗

arXiv2024

FedPEAT: Convergence of Federated Learning, Parameter-Efficient Fine Tuning, and Emulator Assisted Tuning for Artificial Intelligence Foundation Models with Mobile Edge Computing

The emergence of foundation models, including language and vision models, has reshaped AI's landscape, offering capabilities across various applications. Deploying and fine-tuning these large models, like GPT-3 and BERT, presents challenges, especially in the current foundation model era. We introduce Emulator-Assisted Tuning (EAT) combined with Parameter-Efficient Fine-Tuning (PEFT) to form Parameter-Efficient Emulator-Assisted Tuning (PEAT). Further, we expand this into federated learning as Federated PEAT (FedPEAT). FedPEAT uses adapters, emulators, and PEFT for federated model tuning, enhancing model privacy and memory efficiency. Adapters adjust pre-trained models, while emulators give a compact representation of original models, addressing both privacy and efficiency. Adaptable to various neural networks, our approach also uses deep reinforcement learning for hyper-parameter optimization. We tested FedPEAT in a unique scenario with a server participating in collaborative federated tuning, showcasing its potential in tackling foundation model challenges.

cs.LG↗

arXiv2024

Cutting Off the Head Ends the Conflict: A Mechanism for Interpreting and Mitigating Knowledge Conflicts in Language Models

Recently, retrieval augmentation and tool augmentation have demonstrated a remarkable capability to expand the internal memory boundaries of language models (LMs) by providing external context. However, internal memory and external context inevitably clash, leading to knowledge conflicts within LMs. In this paper, we aim to interpret the mechanism of knowledge conflicts through the lens of information flow, and then mitigate conflicts by precise interventions at the pivotal point. We find there are some attention heads with opposite effects in the later layers, where memory heads can recall knowledge from internal memory, and context heads can retrieve knowledge from external context. Moreover, we reveal that the pivotal point at which knowledge conflicts emerge in LMs is the integration of inconsistent information flows by memory heads and context heads. Inspired by the insights, we propose a novel method called Pruning Head via PatH PatcHing (PH3), which can efficiently mitigate knowledge conflicts by pruning conflicting attention heads without updating model parameters. PH3 can flexibly control eight LMs to use internal memory ($\uparrow$ 44.0%) or external context ($\uparrow$ 38.5%). Moreover, PH3 can also improve the performance of LMs on open-domain QA tasks. We also conduct extensive experiments to demonstrate the cross-model, cross-relation, and cross-format generalization of our method.

cs.CL↗

arXiv2024

Advancing Translation Preference Modeling with RLHF: A Step Towards Cost-Effective Solution

Faithfulness, expressiveness, and elegance is the constant pursuit in machine translation. However, traditional metrics like \textit{BLEU} do not strictly align with human preference of translation quality. In this paper, we explore leveraging reinforcement learning with human feedback (\textit{RLHF}) to improve translation quality. It is non-trivial to collect a large high-quality dataset of human comparisons between translations, especially for low-resource languages. To address this issue, we propose a cost-effective preference learning strategy, optimizing reward models by distinguishing between human and machine translations. In this manner, the reward model learns the deficiencies of machine translation compared to human and guides subsequent improvements in machine translation. Experimental results demonstrate that \textit{RLHF} can effectively enhance translation quality and this improvement benefits other translation directions not trained with \textit{RLHF}. Further analysis indicates that the model's language capabilities play a crucial role in preference learning. A reward model with strong language capabilities can more sensitively learn the subtle differences in translation quality and align better with real human translation preferences.

cs.CL↗

arXiv2024

Tug-of-War Between Knowledge: Exploring and Resolving Knowledge Conflicts in Retrieval-Augmented Language Models

Retrieval-augmented language models (RALMs) have demonstrated significant potential in refining and expanding their internal memory by retrieving evidence from external sources. However, RALMs will inevitably encounter knowledge conflicts when integrating their internal memory with external sources. Knowledge conflicts can ensnare RALMs in a tug-of-war between knowledge, limiting their practical applicability. In this paper, we focus on exploring and resolving knowledge conflicts in RALMs. First, we present an evaluation framework for assessing knowledge conflicts across various dimensions. Then, we investigate the behavior and preference of RALMs from the following two perspectives: (1) Conflicts between internal memory and external sources: We find that stronger RALMs emerge with the Dunning-Kruger effect, persistently favoring their faulty internal memory even when correct evidence is provided. Besides, RALMs exhibit an availability bias towards common knowledge; (2) Conflicts between truthful, irrelevant and misleading evidence: We reveal that RALMs follow the principle of majority rule, leaning towards placing trust in evidence that appears more frequently. Moreover, we find that RALMs exhibit confirmation bias, and are more willing to choose evidence that is consistent with their internal memory. To solve the challenge of knowledge conflicts, we propose a method called Conflict-Disentangle Contrastive Decoding (CD2) to better calibrate the model's confidence. Experimental results demonstrate that our CD2 can effectively resolve knowledge conflicts in RALMs.

cs.CL↗

arXiv2024

MoELoRA: Contrastive Learning Guided Mixture of Experts on Parameter-Efficient Fine-Tuning for Large Language Models

Fine-tuning is often necessary to enhance the adaptability of Large Language Models (LLM) to downstream tasks. Nonetheless, the process of updating billions of parameters demands significant computational resources and training time, which poses a substantial obstacle to the widespread application of large-scale models in various scenarios. To address this issue, Parameter-Efficient Fine-Tuning (PEFT) has emerged as a prominent paradigm in recent research. However, current PEFT approaches that employ a limited set of global parameters (such as LoRA, which adds low-rank approximation matrices to all weights) face challenges in flexibly combining different computational modules in downstream tasks. In this work, we introduce a novel PEFT method: MoELoRA. We consider LoRA as Mixture of Experts (MoE), and to mitigate the random routing phenomenon observed in MoE, we propose the utilization of contrastive learning to encourage experts to learn distinct features. We conducted experiments on 11 tasks in math reasoning and common-sense reasoning benchmarks. With the same number of parameters, our approach outperforms LoRA significantly. In math reasoning, MoELoRA achieved an average performance that was 4.2% higher than LoRA, and demonstrated competitive performance compared to the 175B GPT-3.5 on several benchmarks.

cs.CL↗

arXiv2024

ControlLM: Crafting Diverse Personalities for Language Models

As language models continue to scale in size and capability, they display an array of emerging behaviors, both beneficial and concerning. This heightens the need to control model behaviors. We hope to be able to control the personality traits of language models at the inference-time so as to have various character features, on top of which the requirements of different types of tasks can be met. Personality is a higher-level and more abstract behavioral representation for language models. We introduce ControlLM, which leverages differential activation patterns, derived from contrasting behavioral prompts in the model's latent space, to influence the model's personality traits at inference. This approach allows for the precise, real-time adjustment of model behavior. First, we demonstrate ControlLM's capacity to elicit diverse persona behaviors without any training, while precision control allows personality traits to closely match average human values. Subsequently, we showcase improved reasoning and question answering through selective amplification of beneficial attributes like conscientiousness and friendliness. We hope that this work will inspire research on controlling human-like behaviors of language models and provide insights for future research. Our code is publicly available at: https://github.com/wengsyx/ControlLM.

cs.CL↗

arXiv2024

Device Scheduling and Assignment in Hierarchical Federated Learning for Internet of Things

Federated Learning (FL) is a promising machine learning approach for Internet of Things (IoT), but it has to address network congestion problems when the population of IoT devices grows. Hierarchical FL (HFL) alleviates this issue by distributing model aggregation to multiple edge servers. Nevertheless, the challenge of communication overhead remains, especially in scenarios where all IoT devices simultaneously join the training process. For scalability, practical HFL schemes select a subset of IoT devices to participate in the training, hence the notion of device scheduling. In this setting, only selected IoT devices are scheduled to participate in the global training, with each of them being assigned to one edge server. Existing HFL assignment methods are primarily based on search mechanisms, which suffer from high latency in finding the optimal assignment. This paper proposes an improved K-Center algorithm for device scheduling and introduces a deep reinforcement learning-based approach for assigning IoT devices to edge servers. Experiments show that scheduling 50% of IoT devices is generally adequate for achieving convergence in HFL with much lower time delay and energy consumption. In cases where reduction in energy consumption (such as in Green AI) and reduction of messages (to avoid burst traffic) are key objectives, scheduling 30% IoT devices allows a substantial reduction in energy and messages with similar model accuracy.

cs.DC↗

arXiv2024

Deep Learning Innovations in Diagnosing Diabetic Retinopathy: The Potential of Transfer Learning and the DiaCNN Model

Diabetic retinopathy (DR) is a significant cause of vision impairment, emphasizing the critical need for early detection and timely intervention to avert visual deterioration. Diagnosing DR is inherently complex, as it necessitates the meticulous examination of intricate retinal images by experienced specialists. This makes the early diagnosis of DR essential for effective treatment and the prevention of eventual blindness. Traditional diagnostic methods, relying on human interpretation of these medical images, face challenges in terms of accuracy and efficiency. In the present research, we introduce a novel method that offers superior precision in DR diagnosis, compared to these traditional methods, by employing advanced deep learning techniques. Central to this approach is the concept of transfer learning. This entails using pre-existing, well-established models, specifically InceptionResNetv2 and Inceptionv3, to extract features and fine-tune select layers to cater to the unique requirements of this specific diagnostic task. Concurrently, we also present a newly devised model, DiaCNN, which is tailored for the classification of eye diseases. To validate the efficacy of the proposed methodology, we leveraged the Ocular Disease Intelligent Recognition (ODIR) dataset, which comprises eight different eye disease categories. The results were promising. The InceptionResNetv2 model, incorporating transfer learning, registered an impressive 97.5% accuracy in both the training and testing phases. Its counterpart, the Inceptionv3 model, achieved an even more commendable 99.7% accuracy during training, and 97.5% during testing. Remarkably, the DiaCNN model showcased unparalleled precision, achieving 100% accuracy in training and 98.3\% in testing.

eess.IV↗

arXiv2024

Inference for high-dimensional linear expectile regression with de-biased method

In this paper, we address the inference problem in high-dimensional linear expectile regression. We transform the expectile loss into a weighted-least-squares form and apply a de-biased strategy to establish Wald-type tests for multiple constraints within a regularized framework. Simultaneously, we construct an estimator for the pseudo-inverse of the generalized Hessian matrix in high dimension with general amenable regularizers including Lasso and SCAD, and demonstrate its consistency through a new proof technique. We conduct simulation studies and real data applications to demonstrate the efficacy of our proposed test statistic in both homoscedastic and heteroscedastic scenarios.

stat.ME↗

arXiv2024

LLaMA Beyond English: An Empirical Study on Language Capability Transfer

In recent times, substantial advancements have been witnessed in large language models (LLMs), exemplified by ChatGPT, showcasing remarkable proficiency across a range of complex tasks. However, many mainstream LLMs (e.g. LLaMA) are pretrained on English-dominant corpus, which limits their performance in other non-English languages. In this paper, we focus on how to effectively transfer the capabilities of language generation and following instructions to a non-English language. To answer this question, we conduct an extensive empirical investigation based on LLaMA, accumulating over 1440 GPU hours. We analyze the impact of key factors such as vocabulary extension, further pretraining, and instruction tuning on transfer. To accurately assess the model's level of knowledge, we employ four widely used standardized testing benchmarks: C-Eval, MMLU, AGI-Eval, and GAOKAO-Bench. Furthermore, a comprehensive evaluation of the model's response quality is conducted, considering aspects such as accuracy, fluency, informativeness, logical coherence, and harmlessness, based on LLM-Eval, a benchmarks consisting instruction tasks from 17 diverse categories. Our evaluation results demonstrate that comparable performance to state-of-the-art transfer models can be achieved with less than 1% of the pretraining data, both in terms of knowledge alignment and response quality. Furthermore, the experimental outcomes across the thirteen low-resource languages also exhibit similar trends. We anticipate that the conclusions revealed by the experiments will aid the community in developing non-English LLMs.

cs.CL↗

arXiv2024

Secrets of RLHF in Large Language Models Part II: Reward Modeling

Reinforcement Learning from Human Feedback (RLHF) has become a crucial technology for aligning language models with human values and intentions, enabling models to produce more helpful and harmless responses. Reward models are trained as proxies for human preferences to drive reinforcement learning optimization. While reward models are often considered central to achieving high performance, they face the following challenges in practical applications: (1) Incorrect and ambiguous preference pairs in the dataset may hinder the reward model from accurately capturing human intent. (2) Reward models trained on data from a specific distribution often struggle to generalize to examples outside that distribution and are not suitable for iterative RLHF training. In this report, we attempt to address these two issues. (1) From a data perspective, we propose a method to measure the strength of preferences within the data, based on a voting mechanism of multiple reward models. Experimental results confirm that data with varying preference strengths have different impacts on reward model performance. We introduce a series of novel methods to mitigate the influence of incorrect and ambiguous preferences in the dataset and fully leverage high-quality preference data. (2) From an algorithmic standpoint, we introduce contrastive learning to enhance the ability of reward models to distinguish between chosen and rejected responses, thereby improving model generalization. Furthermore, we employ meta-learning to enable the reward model to maintain the ability to differentiate subtle differences in out-of-distribution samples, and this approach can be utilized for iterative RLHF optimization.

cs.AI↗

arXiv2024

A Survey on the Applications of Frontier AI, Foundation Models, and Large Language Models to Intelligent Transportation Systems

This survey paper explores the transformative influence of frontier AI, foundation models, and Large Language Models (LLMs) in the realm of Intelligent Transportation Systems (ITS), emphasizing their integral role in advancing transportation intelligence, optimizing traffic management, and contributing to the realization of smart cities. Frontier AI refers to the forefront of AI technology, encompassing the latest advancements, innovations, and experimental techniques in the field, especially AI foundation models and LLMs. Foundation models, like GPT-4, are large, general-purpose AI models that provide a base for a wide range of applications. They are characterized by their versatility and scalability. LLMs are obtained from finetuning foundation models with a specific focus on processing and generating natural language. They excel in tasks like language understanding, text generation, translation, and summarization. By leveraging vast textual data, including traffic reports and social media interactions, LLMs extract critical insights, fostering the evolution of ITS. The survey navigates the dynamic synergy between LLMs and ITS, delving into applications in traffic management, integration into autonomous vehicles, and their role in shaping smart cities. It provides insights into ongoing research, innovations, and emerging trends, aiming to inspire collaboration at the intersection of language, intelligence, and mobility for safer, more efficient, and sustainable transportation systems. The paper further surveys interactions between LLMs and various aspects of ITS, exploring roles in traffic management, facilitating autonomous vehicles, and contributing to smart city development, while addressing challenges brought by frontier AI and foundation models. This paper offers valuable inspiration for future research and innovation in the transformative domain of intelligent transportation.

cs.CL↗

arXiv2023

Journey to the Center of the Knowledge Neurons: Discoveries of Language-Independent Knowledge Neurons and Degenerate Knowledge Neurons

Pre-trained language models (PLMs) contain vast amounts of factual knowledge, but how the knowledge is stored in the parameters remains unclear. This paper delves into the complex task of understanding how factual knowledge is stored in multilingual PLMs, and introduces the Architecture-adapted Multilingual Integrated Gradients method, which successfully localizes knowledge neurons more precisely compared to current methods, and is more universal across various architectures and languages. Moreover, we conduct an in-depth exploration of knowledge neurons, leading to the following two important discoveries: (1) The discovery of Language-Independent Knowledge Neurons, which store factual knowledge in a form that transcends language. We design cross-lingual knowledge editing experiments, demonstrating that the PLMs can accomplish this task based on language-independent neurons; (2) The discovery of Degenerate Knowledge Neurons, a novel type of neuron showing that different knowledge neurons can store the same fact. Its property of functional overlap endows the PLMs with a robust mastery of factual knowledge. We design fact-checking experiments, proving that the degenerate knowledge neurons can help the PLMs to detect wrong facts. Experiments corroborate these findings, shedding light on the mechanisms of factual knowledge storage in multilingual PLMs, and contribute valuable insights to the field. The code is available at https://github.com/heng840/AMIG.

cs.CL↗

arXiv2023

Human-Centric Resource Allocation in the Metaverse over Wireless Communications

The Metaverse will provide numerous immersive applications for human users, by consolidating technologies like extended reality (XR), video streaming, and cellular networks. Optimizing wireless communications to enable the human-centric Metaverse is important to satisfy the demands of mobile users. In this paper, we formulate the optimization of the system utility-cost ratio (UCR) for the Metaverse over wireless networks. Our human-centric utility measure for virtual reality (VR) applications of the Metaverse represents users' perceptual assessment of the VR video quality as a function of the data rate and the video resolution, and is learnt from real datasets. The variables jointly optimized in our problem include the allocation of both communication and computation resources as well as VR video resolutions. The system cost in our problem comprises the energy consumption and delay, and is non-convex with respect to the optimization variables due to fractions in the mathematical expressions. To solve the non-convex optimization, we develop a novel fractional programming technique, which contributes to optimization theory and has broad applicability beyond our paper. Our proposed algorithm for the system UCR optimization is computationally efficient and finds a stationary point to the constrained optimization. Through extensive simulations, our algorithm is demonstrated to outperform other approaches.

cs.SI↗

arXiv2023

SSTA: Salient Spatially Transformed Attack

Extensive studies have demonstrated that deep neural networks (DNNs) are vulnerable to adversarial attacks, which brings a huge security risk to the further application of DNNs, especially for the AI models developed in the real world. Despite the significant progress that has been made recently, existing attack methods still suffer from the unsatisfactory performance of escaping from being detected by naked human eyes due to the formulation of adversarial example (AE) heavily relying on a noise-adding manner. Such mentioned challenges will significantly increase the risk of exposure and result in an attack to be failed. Therefore, in this paper, we propose the Salient Spatially Transformed Attack (SSTA), a novel framework to craft imperceptible AEs, which enhance the stealthiness of AEs by estimating a smooth spatial transform metric on a most critical area to generate AEs instead of adding external noise to the whole image. Compared to state-of-the-art baselines, extensive experiments indicated that SSTA could effectively improve the imperceptibility of the AEs while maintaining a 100\% attack success rate.

cs.CV↗

arXiv2023

Transforms for Multiplicative and Fractional Programming with Broad Applications in Edge Computing and Communication Networks

Multiplicative Programming (MP) pertains to a spectrum of optimization problems that involve product term(s). As computational paradigms of communication systems continue to evolve, particularly concerning the offloading strategies of computationally intensive tasks simultaneously to centralized or decentralized servers, designing or optimizing effective communication systems with MP techniques becomes increasingly indispensable. Similarly, Fractional Programming (FP) is another significant branch in the optimization domain, addressing various essential scenarios in communication. For instance, in minimization optimization problems, transmission power and processing delay of communication systems are considered critical metrics. In a very recent JSAC paper by Zhao et al. [2], an innovative transform (Zhao's Optimization Transform) was proposed for solving the minimization of MP and FP problems. Nevertheless, the resolution of optimization problems in communication systems encounters several limitations when adopting Zhao's optimization transform, especially in MP problems. Primarily, objective functions proposed in these optimization problems typically involve sum-of-products terms and the optimization variables are always discrete leading to NP-hard problems. Furthermore, multiple functions mapping to the non-negative domain in these scenarios can result in auxiliary variables being zero values, while the same situation is avoidable in FP problems due to the presence of these functions in the denominator. In this paper, we introduce an updated transform, building on the foundations of Zhao's original method, designed to effectively overcome these challenges by reformulating the original problem into a series of convex or concave problems. This introduced problem reformulation provides a superior iteration algorithm with demonstrable convergence to a stationary point.

cs.CC↗

arXiv2023

Mobile Edge Computing and AI Enabled Web3 Metaverse over 6G Wireless Communications: A Deep Reinforcement Learning Approach

The Metaverse is gaining attention among academics as maturing technologies empower the promises and envisagements of a multi-purpose, integrated virtual environment. An interactive and immersive socialization experience between people is one of the promises of the Metaverse. In spite of the rapid advancements in current technologies, the computation required for a smooth, seamless and immersive socialization experience in the Metaverse is overbearing, and the accumulated user experience is essential to be considered. The computation burden calls for computation offloading, where the integration of virtual and physical world scenes is offloaded to an edge server. This paper introduces a novel Quality-of-Service (QoS) model for the accumulated experience in multi-user socialization on a multichannel wireless network. This QoS model utilizes deep reinforcement learning approaches to find the near-optimal channel resource allocation. Comprehensive experiments demonstrate that the adoption of the QoS model enhances the overall socialization experience.

cs.NI↗