Agent-based modeling and simulation has evolved as a powerful tool for modeling complex systems, offering insights into emergent behaviors and interactions among diverse agents. Integrating large language models into agent-based modeling and simulation presents a promising avenue for enhancing simulation capabilities. This paper surveys the landscape of utilizing large language models in agent-based modeling and simulation, examining their challenges and promising future directions. In this survey, since this is an interdisciplinary field, we first introduce the background of agent-based modeling and simulation and large language model-empowered agents. We then discuss the motivation for applying large language models to agent-based simulation and systematically analyze the challenges in environment perception, human alignment, action generation, and evaluation. Most importantly, we provide a comprehensive overview of the recent works of large language model-empowered agent-based modeling and simulation in multiple scenarios, which can be divided into four domains: cyber, physical, social, and hybrid, covering simulation of both real-world and virtual environments. Finally, since this area is new and quickly evolving, we discuss the open problems and promising future directions.
With the prevalence of pre-training-fine-tuning paradigm, how to efficiently adapt the pre-trained model to the downstream tasks has been an intriguing issue. Parameter-Efficient Fine-Tuning (PEFT) methods have been proposed for low-cost adaptation. Although PEFT has demonstrated effectiveness and been widely applied, the underlying principles are still unclear. In this paper, we adopt the PAC-Bayesian generalization error bound, viewing pre-training as a shift of prior distribution which leads to a tighter bound for generalization error. We validate this shift from the perspectives of oscillations in the loss landscape and the quasi-sparsity in gradient distribution. Based on this, we propose a gradient-based sparse fine-tuning algorithm, named Sparse Increment Fine-Tuning (SIFT), and validate its effectiveness on a range of tasks including the GLUE Benchmark and Instruction-tuning. The code is accessible at https://github.com/song-wx/SIFT/.
StarCraft II is a challenging benchmark for AI agents due to the necessity of both precise micro level operations and strategic macro awareness. Previous works, such as Alphastar and SCC, achieve impressive performance on tackling StarCraft II , however, still exhibit deficiencies in long term strategic planning and strategy interpretability. Emerging large language model (LLM) agents, such as Voyage and MetaGPT, presents the immense potential in solving intricate tasks. Motivated by this, we aim to validate the capabilities of LLMs on StarCraft II, a highly complex RTS game.To conveniently take full advantage of LLMs` reasoning abilities, we first develop textual StratCraft II environment, called TextStarCraft II, which LLM agent can interact. Secondly, we propose a Chain of Summarization method, including single frame summarization for processing raw observations and multi frame summarization for analyzing game information, providing command recommendations, and generating strategic decisions. Our experiment consists of two parts: first, an evaluation by human experts, which includes assessing the LLMs`s mastery of StarCraft II knowledge and the performance of LLM agents in the game; second, the in game performance of LLM agents, encompassing aspects like win rate and the impact of Chain of Summarization.Experiment results demonstrate that: 1. LLMs possess the relevant knowledge and complex planning abilities needed to address StarCraft II scenarios; 2. Human experts consider the performance of LLM agents to be close to that of an average player who has played StarCraft II for eight years; 3. LLM agents are capable of defeating the built in AI at the Harder(Lv5) difficulty level. We have open sourced the code and released demo videos of LLM agent playing StarCraft II.
This paper proposes an approach to build 3D scene graphs in arbitrary indoor and outdoor environments. Such extension is challenging; the hierarchy of concepts that describe an outdoor environment is more complex than for indoors, and manually defining such hierarchy is time-consuming and does not scale. Furthermore, the lack of training data prevents the straightforward application of learning-based tools used in indoor settings. To address these challenges, we propose two novel extensions. First, we develop methods to build a spatial ontology defining concepts and relations relevant for indoor and outdoor robot operation. In particular, we use a Large Language Model (LLM) to build such an ontology, thus largely reducing the amount of manual effort required. Second, we leverage the spatial ontology for 3D scene graph construction using Logic Tensor Networks (LTN) to add logical rules, or axioms (e.g., "a beach contains sand"), which provide additional supervisory signals at training time thus reducing the need for labelled data, providing better predictions, and even allowing predicting concepts unseen at training time. We test our approach in a variety of datasets, including indoor, rural, and coastal environments, and show that it leads to a significant increase in the quality of the 3D scene graph generation with sparsely annotated data.
Scalable Vector Graphics (SVGs) are vital for modern image rendering due to their scalability and versatility. Previous SVG generation methods have focused on curve-based vectorization, lacking semantic understanding, often producing artifacts, and struggling with SVG primitives beyond path curves. To address these issues, we introduce StarVector, a multimodal large language model for SVG generation. It performs image vectorization by understanding image semantics and using SVG primitives for compact, precise outputs. Unlike traditional methods, StarVector works directly in the SVG code space, leveraging visual understanding to apply accurate SVG primitives. To train StarVector, we create SVG-Stack, a diverse dataset of 2M samples that enables generalization across vectorization tasks and precise use of primitives like ellipses, polygons, and text. We address challenges in SVG evaluation, showing that pixel-based metrics like MSE fail to capture the unique qualities of vector graphics. We introduce SVG-Bench, a benchmark across 10 datasets, and 3 tasks: Image-to-SVG, Text-to-SVG generation, and diagram generation. Using this setup, StarVector achieves state-of-the-art performance, producing more compact and semantically rich SVGs.
This research introduces an innovative evaluation method for the "founder-idea" fit in early-stage startups, utilizing advanced large language model techniques to assess founders' profiles against their startup ideas to enhance decision-making. Embeddings, self-play, tree-of-thought, and critique-based refinement techniques show early promising results that each idea's success patterns are unique and they should be evaluated based on the context of the founder's background.
Introduced to enhance the efficiency of large language model (LLM) inference, speculative decoding operates by having a smaller model generate a draft. A larger target model then reviews this draft to align with its output, and any acceptance by the target model results in a reduction of the number of the target model runs, ultimately improving efficiency. However, the drafting process in speculative decoding includes slow autoregressive generation and allocates equal time to generating tokens, irrespective of their importance. These inefficiencies collectively contribute to the suboptimal performance of speculative decoding. To further improve LLM inference, we introduce Cascade Speculative Drafting (CS Drafting), a speculative execution algorithm that incorporates two types of cascades. The Vertical Cascade eliminates autoregressive generation from neural models, while the Horizontal Cascade optimizes time allocation in drafting for improved efficiency. Combining both cascades, CS Drafting achieves up to an 81 percent additional speedup over speculative decoding in our experiments, while maintaining the same output distribution as the target model. Our code is publicly available at https://github.com/lfsszd/CS-Drafting.
Large language models (LLMs) have shown remarkable proficiency in human-level reasoning and generation capabilities, which encourages extensive research on their application in mathematical problem solving. However, current work has been largely focused on text-based mathematical problems, with limited investigation in problems involving geometric information. Addressing this gap, we aim to enable LLMs to solve geometric problems by understanding image input. We first analyze the limitations of current Multimodal Large Language Models (MLLMs) in this area: they struggle to accurately comprehending basic geometric elements and their relationships. To overcome these challenges, we take advantage of the unique characteristics of geometric problems (such as unique geometric logical form, and geometric scalability) and the capacity of the textual LLMs to build an enriched multimodal geometry dataset based on existing data. The augmented dataset, Geo170K, contains more than 170K geometric image-caption and question-answer pairs. Utilizing our constructed Geo170K dataset, we develop G-LLaVA, which demonstrates exceptional performance in solving geometric problems, significantly outperforming GPT-4-V on the MathVista benchmark with only 7B parameters.
Retrieval-Augmented Generation (RAG) grounds Large Language Model (LLM) output by leveraging external knowledge sources to reduce factual hallucinations. However, prior work lacks a comprehensive evaluation of different language families, making it challenging to evaluate LLM robustness against errors in external retrieved knowledge. To overcome this, we establish NoMIRACL, a human-annotated dataset for evaluating LLM robustness in RAG across 18 typologically diverse languages. NoMIRACL includes both a non-relevant and a relevant subset. Queries in the non-relevant subset contain passages judged as non-relevant, whereas queries in the relevant subset include at least a single judged relevant passage. We measure relevance assessment using: (i) hallucination rate, measuring model tendency to hallucinate, when the answer is not present in passages in the non-relevant subset, and (ii) error rate, measuring model inaccuracy to recognize relevant passages in the relevant subset.In our work, we observe that most models struggle to balance the two capacities. Models such as LLAMA-2 and Orca-2 achieve over 88% hallucination rate on the non-relevant subset. Mistral and LLAMA-3 hallucinate less but can achieve up to a 74.9% error rate on the relevant subset. Overall, GPT-4 is observed to provide the best tradeoff on both subsets, highlighting future work necessary to improve LLM robustness. NoMIRACL dataset and evaluation code are available at: https://github.com/project-miracl/nomiracl.
In this technical report, we present VinaLLaMA, an open-weight, state-of-the-art (SOTA) Large Language Model for the Vietnamese language, built upon LLaMA-2 with an additional 800 billion trained tokens. VinaLLaMA not only demonstrates fluency in Vietnamese but also exhibits a profound understanding of Vietnamese culture, making it a truly indigenous model. VinaLLaMA-7B-chat, trained on 1 million high-quality synthetic samples, achieves SOTA results on key benchmarks, including VLSP, VMLU, and Vicuna Benchmark Vietnamese, marking a significant advancement in the Vietnamese AI landscape and offering a versatile resource for various applications.
This paper studies the alignment process of generative models with Reinforcement Learning from Human Feedback (RLHF). We first identify the primary challenges of existing popular methods like offline PPO and offline DPO as lacking in strategical exploration of the environment. Then, to understand the mathematical principle of RLHF, we consider a standard mathematical formulation, the reverse-KL regularized contextual bandit for RLHF. Despite its widespread practical application, a rigorous theoretical analysis of this formulation remains open. We investigate its behavior in three distinct settings -- offline, online, and hybrid -- and propose efficient algorithms with finite-sample theoretical guarantees. Moving towards practical applications, our framework, with a robust approximation of the information-theoretical policy improvement oracle, naturally gives rise to several novel RLHF algorithms. This includes an iterative version of the Direct Preference Optimization (DPO) algorithm for online settings, and a multi-step rejection sampling strategy for offline scenarios. Our empirical evaluations on real-world alignment experiment of large language model demonstrate that these proposed methods significantly surpass existing strong baselines, such as DPO and Rejection Sampling Optimization (RSO), showcasing the connections between solid theoretical foundations and their potent practical implementations.
Click-Through Rate (CTR) prediction is a crucial task in online recommendation platforms as it involves estimating the probability of user engagement with advertisements or items by clicking on them. Given the availability of various services like online shopping, ride-sharing, food delivery, and professional services on commercial platforms, recommendation systems in these platforms are required to make CTR predictions across multiple domains rather than just a single domain. However, multi-domain click-through rate (MDCTR) prediction remains a challenging task in online recommendation due to the complex mutual influence between domains. Traditional MDCTR models typically encode domains as discrete identifiers, ignoring rich semantic information underlying. Consequently, they can hardly generalize to new domains. Besides, existing models can be easily dominated by some specific domains, which results in significant performance drops in the other domains (i.e. the "seesaw phenomenon"). In this paper, we propose a novel solution Uni-CTR to address the above challenges. Uni-CTR leverages a backbone Large Language Model (LLM) to learn layer-wise semantic representations that capture commonalities between domains. Uni-CTR also uses several domain-specific networks to capture the characteristics of each domain. Note that we design a masked loss strategy so that these domain-specific networks are decoupled from backbone LLM. This allows domain-specific networks to remain unchanged when incorporating new or removing domains, thereby enhancing the flexibility and scalability of the system significantly. Experimental results on three public datasets show that Uni-CTR outperforms the state-of-the-art (SOTA) MDCTR models significantly. Furthermore, Uni-CTR demonstrates remarkable effectiveness in zero-shot prediction. We have applied Uni-CTR in industrial scenarios, confirming its efficiency.
In the realm of public health, vaccination stands as the cornerstone for mitigating disease risks and controlling their proliferation. The recent COVID-19 pandemic has highlighted how vaccines play a crucial role in keeping us safe. However the situation involves a mix of perspectives, with skepticism towards vaccines prevailing for various reasons such as political dynamics, apprehensions about side effects, and more. The paper addresses the challenge of comprehensively understanding and categorizing these diverse concerns expressed in the context of vaccination. Our focus is on developing a robust multi-label classifier capable of assigning specific concern labels to tweets based on the articulated apprehensions towards vaccines. To achieve this, we delve into the application of a diverse set of advanced natural language processing techniques and machine learning algorithms including transformer models like BERT, state of the art GPT 3.5, Classifier Chains & traditional methods like SVM, Random Forest, Naive Bayes. We see that the cutting-edge large language model outperforms all other methods in this context.
Knowledge distillation (KD) has been widely employed to transfer knowledge from a large language model (LLM) to a specialized model in low-data regimes through pseudo label learning. However, pseudo labels generated by teacher models are usually noisy and may influence KD performance. This study delves into KD with noisy teachers and uncovers that the student model can already generate more accurate predictions than the teacher labels used to train it during KD, indicating its inherent ability to denoise noisy teacher labels. Motivated by this finding, we propose Peer-Advised KD to improve vanilla KD from noisy teachers. Experiments show that Peer-Advised KD can outperform LLM by approximately 5% with 50 human-labeled data, and even competitive to standard supervised finetuning with 750 human-labeled data.
Instruction tuning significantly enhances the performance of large language models (LLMs) across various tasks. However, the procedure to optimizing the mixing of instruction datasets for LLM fine-tuning is still poorly understood. This study categorizes instructions into three primary types: NLP downstream tasks, coding, and general chat. We explore the effects of instruction tuning on different combinations of datasets on LLM performance, and find that certain instruction types are more advantageous for specific applications but can negatively impact other areas. This work provides insights into instruction mixtures, laying the foundations for future research.
To understand how well a large language model captures certain semantic or syntactic features, researchers typically apply probing classifiers. However, the accuracy of these classifiers is critical for the correct interpretation of the results. If a probing classifier exhibits low accuracy, this may be due either to the fact that the language model does not capture the property under investigation, or to shortcomings in the classifier itself, which is unable to adequately capture the characteristics encoded in the internal representations of the model. Consequently, for more effective diagnosis, it is necessary to use the most accurate classifiers possible for a particular type of task. Logistic regression on the output representation of the transformer neural network layer is most often used to probing the syntactic properties of the language model. We show that using gradient boosting decision trees at the Knowledge Neuron layer, i.e., at the hidden layer of the feed-forward network of the transformer as a probing classifier for recognizing parts of a sentence is more advantageous than using logistic regression on the output representations of the transformer layer. This approach is also preferable to many other methods. The gain in error rate, depending on the preset, ranges from 9-54%
Beyond 5G networks provide solutions for next-generation communications, especially digital twins networks (DTNs) have gained increasing popularity for bridging physical space and digital space. However, current DTNs networking frameworks pose a number of challenges especially when applied in scenarios that require high communication efficiency and multimodal data processing. First, current DTNs frameworks are unavoidable regarding high resource consumption and communication congestion because of original bit-level communication and high-frequency computation, especially distributed learning-based DTNs. Second, current machine learning models for DTNs are domain-specific (e.g. E-health), making it difficult to handle DT scenarios with multimodal data processing requirements. Last but not least, current security schemes for DTNs, such as blockchain, introduce additional overheads that impair the efficiency of DTNs. To address the above challenges, we propose a large language model (LLM) empowered DTNs networking framework, LLM-Twin. First, we design the mini-giant model collaboration scheme to achieve efficient deployment of LLM in DTNs, since LLM are naturally conducive to processing multimodal data. Then, we design a semantic-level high-efficiency, and secure communication model for DTNs. The feasibility of LLM-Twin is demonstrated by numerical experiments and case studies. To our knowledge, this is the first to propose LLM-based semantic-level digital twin networking framework.
Crash bugs cause unexpected program behaviors or even termination, requiring high-priority resolution. However, manually resolving crash bugs is challenging and labor-intensive, and researchers have proposed various techniques for their automated localization and repair. ChatGPT, a recent large language model (LLM), has garnered significant attention due to its exceptional performance across various domains. This work performs the first investigation into ChatGPT's capability in resolve real-world crash bugs, focusing on its effectiveness in both localizing and repairing code-related and environment-related crash bugs. Specifically, we initially assess ChatGPT's fundamental ability to resolve crash bugs with basic prompts in a single iteration. We observe that ChatGPT performs better at resolving code-related crash bugs compared to environment-related ones, and its primary challenge in resolution lies in inaccurate localization. Additionally, we explore ChatGPT's potential with various advanced prompts. Furthermore, by stimulating ChatGPT's self-planning, it methodically investigates each potential crash-causing environmental factor through proactive inquiry, ultimately identifying the root cause of the crash. Based on our findings, we propose IntDiagSolver, an interaction methodology designed to facilitate precise crash bug resolution through continuous interaction with LLMs. Evaluating IntDiagSolver on multiple LLMs reveals consistent enhancement in the accuracy of crash bug resolution, including ChatGPT, Claude, and CodeLlama.
Multimodal knowledge bases (MMKBs) provide cross-modal aligned knowledge crucial for multimodal tasks. However, the images in existing MMKBs are generally collected for entities in encyclopedia knowledge graphs. Therefore, detailed groundings of visual semantics with linguistic concepts are lacking, which are essential for the visual concept cognition ability of multimodal models. Addressing this gap, we introduce M^2ConceptBase, the first concept-centric MMKB. M^2ConceptBase models concepts as nodes with associated images and detailed textual descriptions. We propose a context-aware multimodal symbol grounding approach to align concept-image and concept-description pairs using context information from image-text datasets. Comprising 951K images and 152K concepts, M^2ConceptBase links each concept to an average of 6.27 images and a single description, ensuring comprehensive visual and textual semantics. Human studies confirm more than 95% alignment accuracy, underscoring its quality. Additionally, our experiments demonstrate that M^2ConceptBase significantly enhances VQA model performance on the OK-VQA task. M^2ConceptBase also substantially improves the fine-grained concept understanding capabilities of multimodal large language models through retrieval augmentation in two concept-related tasks, highlighting its value.
This paper explores the image synthesis capabilities of GPT-4, a leading multi-modal large language model. We establish a benchmark for evaluating the fidelity of texture features in images generated by GPT-4, comprising manually painted pictures and their AI-generated counterparts. The contributions of this study are threefold: First, we provide an in-depth analysis of the fidelity of image synthesis features based on GPT-4, marking the first such study on this state-of-the-art model. Second, the quantitative and qualitative experiments fully reveals the limitations of the GPT-4 model in image synthesis. Third, we have compiled a unique benchmark of manual drawings and corresponding GPT-4-generated images, introducing a new task to advance fidelity research in AI-generated content (AIGC). The dataset is available at: \url{https://github.com/rickwang28574/DeepArt}.