Search

Large language models are rapidly gaining popularity and have been widely adopted in real-world applications. While the quality of training data is essential, privacy concerns arise during data collection. Federated learning offers a solution by allowing multiple clients to collaboratively train LLMs without sharing local data. However, FL introduces new challenges, such as model convergence issues due to heterogeneous data and high communication costs. A comprehensive study is required to address these challenges and guide future research. This paper surveys Federated learning for LLMs (FedLLM), highlighting recent advances and future directions. We focus on two key aspects: fine-tuning and prompt learning in a federated setting, discussing existing work and associated research challenges. We finally propose potential research directions for federated LLMs, including pre-training and how LLMs can further enhance federated learning.

The purpose of face super-resolution (FSR) is to reconstruct high-resolution (HR) face images from low-resolution (LR) inputs. With the continuous advancement of deep learning technologies, contemporary prior-guided FSR methods initially estimate facial priors and then use this information to assist in the super-resolution reconstruction process. However, ensuring the accuracy of prior estimation remains challenging, and straightforward cascading and convolutional operations often fail to fully leverage prior knowledge. Inaccurate or insufficiently utilized prior information inevitably degrades FSR performance. To address this issue, we propose a prior knowledge distillation network (PKDN) for FSR, which involves transferring prior information from the teacher network to the student network. This approach enables the network to learn priors during the training stage while relying solely on low-resolution facial images during the testing stage, thus mitigating the adverse effects of prior estimation inaccuracies. Additionally, we incorporate robust attention mechanisms to design a parsing map fusion block that effectively utilizes prior information. To prevent feature loss, we retain multi-scale features during the feature extraction stage and employ them in the subsequent super-resolution reconstruction process. Experimental results on benchmark datasets demonstrate that our PKDN approach surpasses existing FSR methods in generating high-quality face images.

Large Language Models (LLMs) are powerful tools with the potential to benefit society immensely, yet, they have demonstrated biases that perpetuate societal inequalities. Despite significant advancements in bias mitigation techniques using data augmentation, zero-shot prompting, and model fine-tuning, biases continuously persist, including subtle biases that may elude human detection. Recent research has shown a growing interest in multi-LLM approaches, which have been demonstrated to be effective in improving the quality of reasoning and factuality in LLMs. Building on this approach, we propose a novel multi-LLM debiasing framework aimed at reducing bias in LLMs. Our work is the first to introduce and evaluate two distinct approaches within this framework for debiasing LLMs: a centralized method, where the conversation is facilitated by a single central LLM, and a decentralized method, where all models communicate directly. Our findings reveal that our multi-LLM framework significantly reduces bias in LLMs, outperforming the baseline method across several social groups.

Multimodal large language models (MLLMs) equip pre-trained large-language models (LLMs) with visual capabilities. While textual prompting in LLMs has been widely studied, visual prompting has emerged for more fine-grained and free-form visual instructions. This paper presents the first comprehensive survey on visual prompting methods in MLLMs, focusing on visual prompting, prompt generation, compositional reasoning, and prompt learning. We categorize existing visual prompts and discuss generative methods for automatic prompt annotations on the images. We also examine visual prompting methods that enable better alignment between visual encoders and backbone LLMs, concerning MLLM's visual grounding, object referring, and compositional reasoning abilities. In addition, we provide a summary of model training and in-context learning methods to improve MLLM's perception and understanding of visual prompts. This paper examines visual prompting methods developed in MLLMs and provides a vision of the future of these methods., Comment: 10 pages

The retrieval augmented generation (RAG) framework addresses an ambiguity in user queries in QA systems by retrieving passages that cover all plausible interpretations and generating comprehensive responses based on the passages. However, our preliminary studies reveal that a single retrieval process often suffers from low quality results, as the retrieved passages frequently fail to capture all plausible interpretations. Although the iterative RAG approach has been proposed to address this problem, it comes at the cost of significantly reduced efficiency. To address these issues, we propose the diversify-verify-adapt (DIVA) framework. DIVA first diversifies the retrieved passages to encompass diverse interpretations. Subsequently, DIVA verifies the quality of the passages and adapts the most suitable approach tailored to their quality. This approach improves the QA systems accuracy and robustness by handling low quality retrieval issue in ambiguous questions, while enhancing efficiency.

Relativistic spin-polarized electron beams are important for fundamental research and the industry, but their generation currently requires conventional accelerators or ultrastrong laser facilities, limiting their accessibility and broad applications. Here, we put forward a novel method for constructing a compact efficient "polarizer" that achieves direct ultrafast conversion of relativistic dense electron beams into polarized ones, based on the beam "self-polarization" mechanism via simple beam-target interactions. In this scheme, as the electron beam grazes through the polarizer (a double-layer solid target), it ionizes the target and excites an asymmetric plasma field due to the plasma backflows. This field then reacts on the beam itself, triggering spontaneous radiative polarization and reflection of the beam, and ultimately yielding a dense polarized electron beam. Moreover, the double-layer target setup induces a plasma bubble that focuses the polarized beam and reshapes its polarization distribution. Our method is robust with respect to the beam and target parameters, and opens a new avenue for relativistic beam polarization with compact accessible devices, which would facilitate their broad applications and the development of related experiments, such as in strong-field QED studies, and polarized electron-positron and electron-ion colliders.

Material responses to static and dynamic stimuli, represented as nonlinear curves, are design targets for engineering functionalities like structural support, impact protection, and acoustic and photonic bandgaps. Three-dimensional metamaterials offer significant tunability due to their internal structure, yet existing methods struggle to capture their complex behavior-to-structure relationships. We present GraphMetaMat, a graph-based framework capable of designing three-dimensional metamaterials with programmable responses and arbitrary manufacturing constraints. Integrating graph networks, physics biases, reinforcement learning, and tree search, GraphMetaMat can target stress-strain curves spanning four orders of magnitude and complex behaviors, as well as viscoelastic transmission responses with varying attenuation gaps. GraphMetaMat can create cushioning materials for protective equipment and vibration-damping panels for electric vehicles, outperforming commercial materials, and enabling the automatic design of materials with on-demand functionalities., Comment: 19 pages, 5 figures

The extended Berkeley Packet Filter (eBPF) is extensively utilized for observability and performance analysis in cloud-native environments. However, deploying eBPF programs across a heterogeneous cloud environment presents challenges, including compatibility issues across different kernel versions, operating systems, runtimes, and architectures. Traditional deployment methods, such as standalone containers or tightly integrated core applications, are cumbersome and inefficient, particularly when dynamic plugin management is required. To address these challenges, we introduce Wasm-bpf, a lightweight runtime on WebAssembly and the WebAssembly System Interface (WASI). Leveraging Wasm platform independence and WASI standardized system interface, with enhanced relocation for different architectures, Wasm-bpf ensures cross-platform compatibility for eBPF programs. It simplifies deployment by integrating with container toolchains, allowing eBPF programs to be packaged as Wasm modules that can be easily managed within cloud environments. Additionally, Wasm-bpf supports dynamic plugin management in WebAssembly. Our implementation and evaluation demonstrate that Wasm-bpf introduces minimal overhead compared to native eBPF implementations while simplifying the deployment process.

Large language models (LLMs) have been applied to a wide range of tasks, including text summarization, web navigation, and chatbots. They have benefitted from supervised fine-tuning (SFT) and reinforcement learning from human feedback (RLHF) following an unsupervised pretraining. These datasets can be difficult to collect, limited in scope, and vary in sample quality. Additionally, datasets can vary extensively in supervision format, from numerical to binary as well as multi-dimensional with many different values. We present a framework for fine-tuning LLMs using heterogeneous feedback, which has two main components. First, we combine the heterogeneous feedback data into a single supervision format, compatible with methods like SFT and RLHF. Next, given this unified feedback dataset, we extract a high-quality and diverse subset to obtain performance increases potentially exceeding the full dataset. We conduct extensive experiments to understand the effectiveness of these techniques for incorporating heterogeneous feedback, and demonstrate improvements from using a high-quality and diverse subset of the data. We find that our framework is able to improve models in multiple areas simultaneously, such as in instruction following and bias reduction., Comment: 7 pages, 1 figure

Instruction tuning in multimodal large language models (MLLMs) aims to smoothly integrate a backbone LLM with a pre-trained feature encoder for downstream tasks. The major challenge is how to efficiently find the synergy through cooperative learning where LLMs adapt their reasoning abilities in downstream tasks while feature encoders adjust their encoding to provide more relevant modal information. In this paper, we analyze the MLLM instruction tuning from both theoretical and empirical perspectives, where we find unbalanced learning between the two components, i.e., the feature encoder and the LLM, can cause diminishing learning gradients that slow the model convergence and often lead to sub-optimal results due to insufficient learning. Inspired by our findings, we propose a measurement to quantitatively evaluate the learning balance, based on which we further design a dynamic learning scheduler that better coordinates the learning. In addition, we introduce an auxiliary loss regularization method to promote updating of the generation distribution of MLLMs considering the learning state of each model component, which potentially prevents each component from gradient diminishing and enables a more accurate estimation of the learning balance coefficient. We conduct experiments with multiple LLM backbones and feature encoders, where our techniques are model-agnostic and can be generically integrated with various MLLM backbones. Experiment results on multiple downstream tasks and modalities in vision and audio, demonstrate the proposed method's better efficiency and effectiveness in MLLM instruction tuning., Comment: 9 pages

Parameter extraction of photovoltaic (PV) models is crucial for the planning, optimization, and control of PV systems. Although some methods using meta-heuristic algorithms have been proposed to determine these parameters, the robustness of solutions obtained by these methods faces great challenges when the complexity of the PV model increases. The unstable results will affect the reliable operation and maintenance strategies of PV systems. In response to this challenge, an improved RIME algorithm with enhanced exploration and exploitation is proposed for robust and accurate parameter identification for various PV models. Specifically, the differential evolution mutation operator is integrated in the exploration phase to enhance the population diversity. Meanwhile, a new exploitation strategy incorporating randomization and neighborhood strategies simultaneously is developed to maintain the balance of exploitation width and depth. The improved RIME algorithm is applied to estimate the optimal parameters of the single-diode model (SDM), double-diode model (DDM), and triple-diode model (TDM) combined with the Lambert-W function for three PV cell and module types including RTC France, Photo Watt-PWP 201 and S75. According to the statistical analysis in 100 runs, the TEIMRE achieves more accurate and robust parameter estimations than other techniques to various PV models in varying environmental conditions. All of our source codes are publicly available at https://github.com/dirge1/TERIME.

Discovering causal relations from observational time series without making the stationary assumption is a significant challenge. In practice, this challenge is common in many areas, such as retail sales, transportation systems, and medical science. Here, we consider this problem for a class of non-stationary time series. The structural causal model (SCM) of this type of time series, called the semi-stationary time series, exhibits that a finite number of different causal mechanisms occur sequentially and periodically across time. This model holds considerable practical utility because it can represent periodicity, including common occurrences such as seasonality and diurnal variation. We propose a constraint-based, non-parametric algorithm for discovering causal relations in this setting. The resulting algorithm, PCMCI$_{\Omega}$, can capture the alternating and recurring changes in the causal mechanisms and then identify the underlying causal graph with conditional independence (CI) tests. We show that this algorithm is sound in identifying causal relations on discrete time series. We validate the algorithm with extensive experiments on continuous and discrete simulated data. We also apply our algorithm to a real-world climate dataset.

Change point detection in time series seeks to identify times when the probability distribution of time series changes. It is widely applied in many areas, such as human-activity sensing and medical science. In the context of multivariate time series, this typically involves examining the joint distribution of high-dimensional data: If any one variable changes, the whole time series is assumed to have changed. However, in practical applications, we may be interested only in certain components of the time series, exploring abrupt changes in their distributions in the presence of other time series. Here, assuming an underlying structural causal model that governs the time-series data generation, we address this problem by proposing a two-stage non-parametric algorithm that first learns parts of the causal structure through constraint-based discovery methods. The algorithm then uses conditional relative Pearson divergence estimation to identify the change points. The conditional relative Pearson divergence quantifies the distribution disparity between consecutive segments in the time series, while the causal discovery method enables a focus on the causal mechanism, facilitating access to independent and identically distributed (IID) samples. Theoretically, the typical assumption of samples being IID in conventional change point detection methods can be relaxed based on the Causal Markov Condition. Through experiments on both synthetic and real-world datasets, we validate the correctness and utility of our approach.

Large Language Models (LLMs) have been achieving competent performance on a wide range of downstream tasks, yet existing work shows that inference on structured data is challenging for LLMs. This is because LLMs need to either understand long structured data or select the most relevant evidence before inference, and both approaches are not trivial. This paper proposes a framework, Learning to Reduce, that fine-tunes a language model with On-Policy Learning to generate a reduced version of an input structured data. When compared to state-of-the-art LLMs like GPT-4, Learning to Reduce not only achieves outstanding performance in reducing the input, but shows generalizability on different datasets. We further show that the model fine-tuned with our framework helps LLMs better perform on table QA tasks especially when the context is longer., Comment: ICML 2024 Workshop on Long-Context Foundation Models, Vienna, Austria 2024. arXiv admin note: substantial text overlap with arXiv:2402.14195

Diffusion models have demonstrated exceptional capabilities in generating a broad spectrum of visual content, yet their proficiency in rendering text is still limited: they often generate inaccurate characters or words that fail to blend well with the underlying image. To address these shortcomings, we introduce a new framework named ARTIST. This framework incorporates a dedicated textual diffusion model to specifically focus on the learning of text structures. Initially, we pretrain this textual model to capture the intricacies of text representation. Subsequently, we finetune a visual diffusion model, enabling it to assimilate textual structure information from the pretrained textual model. This disentangled architecture design and the training strategy significantly enhance the text rendering ability of the diffusion models for text-rich image generation. Additionally, we leverage the capabilities of pretrained large language models to better interpret user intentions, contributing to improved generation quality. Empirical results on the MARIO-Eval benchmark underscore the effectiveness of the proposed method, showing an improvement of up to 15\% in various metrics.

Yuan 2.0-M32, with a similar base architecture as Yuan-2.0 2B, uses a mixture-of-experts architecture with 32 experts of which 2 experts are active. A new router network, Attention Router, is proposed and adopted for a more efficient selection of experts, which improves the accuracy compared to the model with classical router network. Yuan 2.0-M32 is trained with 2000B tokens from scratch, and the training computation consumption is only 9.25% of a dense model at the same parameter scale. Yuan 2.0-M32 demonstrates competitive capability on coding, math, and various domains of expertise, with only 3.7B active parameters of 40B in total, and 7.4 GFlops forward computation per token, both of which are only 1/19 of Llama3-70B. Yuan 2.0-M32 surpass Llama3-70B on MATH and ARC-Challenge benchmark, with accuracy of 55.89 and 95.8 respectively. The models and source codes of Yuan 2.0-M32 are released at Github1., Comment: 14 pages,3 figures, 7 tables

Large language models (LLMs) have demonstrated impressive capabilities in various reasoning tasks, aided by techniques like chain-of-thought (CoT) prompting that elicits verbalized reasoning. However, LLMs often generate text with obvious mistakes and contradictions, raising doubts about their ability to robustly process and utilize generated rationales. In this work, we investigate CoT reasoning in LLMs through the lens of internal representations, focusing on how these representations are influenced by generated rationales. Our preliminary analysis reveals that while generated rationales improve answer accuracy, inconsistencies emerge between the model's internal representations in middle layers and those in final layers, potentially undermining the reliability of their reasoning processes. To address this, we propose internal consistency as a measure of the model's confidence by examining the agreement of latent predictions decoded from intermediate layers. Extensive empirical studies across different models and datasets demonstrate that internal consistency effectively distinguishes between correct and incorrect reasoning paths. Motivated by this, we propose a new approach to calibrate CoT reasoning by up-weighting reasoning paths with high internal consistency, resulting in a significant boost in reasoning performance. Further analysis uncovers distinct patterns in attention and feed-forward modules across layers, providing insights into the emergence of internal inconsistency. In summary, our results demonstrate the potential of using internal representations for self-evaluation of LLMs.

Self-supervised pretraining (SSP) has been recognized as a method to enhance prediction accuracy in various downstream tasks. However, its efficacy for DNA sequences remains somewhat constrained. This limitation stems primarily from the fact that most existing SSP approaches in genomics focus on masked language modeling of individual sequences, neglecting the crucial aspect of encoding statistics across multiple sequences. To overcome this challenge, we introduce an innovative deep neural network model, which incorporates collaborative learning between a `student' and a `teacher' subnetwork. In this model, the student subnetwork employs masked learning on nucleotides and progressively adapts its parameters to the teacher subnetwork through an exponential moving average approach. Concurrently, both subnetworks engage in contrastive learning, deriving insights from two augmented representations of the input sequences. This self-distillation process enables our model to effectively assimilate both contextual information from individual sequences and distributional data across the sequence population. We validated our approach with preliminary pretraining using the human reference genome, followed by applying it to 20 downstream inference tasks. The empirical results from these experiments demonstrate that our novel method significantly boosts inference performance across the majority of these tasks. Our code is available at https://github.com/wiedersehne/FinDNA.

Chain-of-Thought (CoT) has been a widely adopted prompting method, eliciting impressive reasoning abilities of Large Language Models (LLMs). Inspired by the sequential thought structure of CoT, a number of Chain-of-X (CoX) methods have been developed to address various challenges across diverse domains and tasks involving LLMs. In this paper, we provide a comprehensive survey of Chain-of-X methods for LLMs in different contexts. Specifically, we categorize them by taxonomies of nodes, i.e., the X in CoX, and application tasks. We also discuss the findings and implications of existing CoX methods, as well as potential future directions. Our survey aims to serve as a detailed and up-to-date resource for researchers seeking to apply the idea of CoT to broader scenarios.

Nonlinear Compton scattering driven by ultraintense lasers presents a promising avenue for enhancing the photon energy, brilliance, and setup compactness of $\gamma$-ray sources. However, a significant challenge lies in achieving a high polarization degree with commonly generated unpolarized electrons, thus addressing a longstanding puzzle in the field. Here we investigate the polarization dynamics of photons emitted by an unpolarized electron beam interacting with a counter-propagating ultraintense laser pulse numerically, and propose a novel method to generate highly polarized $\gamma$ rays via nonlinear Compton scattering with the aid of vacuum dichroism effect. Our simulations reveal that high-brilliance $\gamma$ rays with polarization beyond 90\% are feasible in a single-shot interaction, rivaling the highest achieved by any $\gamma$-ray sources to date, based on a developed Monte Carlo method incorporating polarization-resolved tree processes of nonlinear Compton scattering and Breit-Wheeler pair production and one-loop vacuum polarization. This generation method showcases an extraordinary ultra-high polarization degree and a user-friendly all-optical experimental setup, while harnessing the high photon energy and brilliance characteristic of nonlinear Compton scattering sources, thus making it of great potential for experimental applications.

Retrieval-augmented large language models (LLMs) have been remarkably competent in various NLP tasks. Despite their great success, the knowledge provided by the retrieval process is not always useful for improving the model prediction, since in some samples LLMs may already be quite knowledgeable and thus be able to answer the question correctly without retrieval. Aiming to save the cost of retrieval, previous work has proposed to determine when to do/skip the retrieval in a data-aware manner by analyzing the LLMs' pretraining data. However, these data-aware methods pose privacy risks and memory limitations, especially when requiring access to sensitive or extensive pretraining data. Moreover, these methods offer limited adaptability under fine-tuning or continual learning settings. We hypothesize that token embeddings are able to capture the model's intrinsic knowledge, which offers a safer and more straightforward way to judge the need for retrieval without the privacy risks associated with accessing pre-training data. Moreover, it alleviates the need to retain all the data utilized during model pre-training, necessitating only the upkeep of the token embeddings. Extensive experiments and in-depth analyses demonstrate the superiority of our model-aware approach.

Large Language Models (LLMs) have shown propensity to generate hallucinated outputs, i.e., texts that are factually incorrect or unsupported. Existing methods for alleviating hallucinations typically require costly human annotations to identify and correct hallucinations in LLM outputs. Moreover, most of these methods focus on a specific type of hallucination, e.g., entity or token errors, which limits their effectiveness in addressing various types of hallucinations exhibited in LLM outputs. To our best knowledge, in this paper we propose the first active learning framework to alleviate LLM hallucinations, reducing costly human annotations of hallucination needed. By measuring fine-grained hallucinations from errors in semantic frame, discourse and content verifiability in text summarization, we propose HAllucination Diversity-Aware Sampling (HADAS) to select diverse hallucinations for annotations in active learning for LLM finetuning. Extensive experiments on three datasets and different backbone models demonstrate advantages of our method in effectively and efficiently mitigating LLM hallucinations., Comment: Accepted to NAACL 2024

Crafting effective captions for figures is important. Readers heavily depend on these captions to grasp the figure's message. However, despite a well-developed set of AI technologies for figures and captions, these have rarely been tested for usefulness in aiding caption writing. This paper introduces SciCapenter, an interactive system that puts together cutting-edge AI technologies for scientific figure captions to aid caption composition. SciCapenter generates a variety of captions for each figure in a scholarly article, providing scores and a comprehensive checklist to assess caption quality across multiple critical aspects, such as helpfulness, OCR mention, key takeaways, and visual properties reference. Users can directly edit captions in SciCapenter, resubmit for revised evaluations, and iteratively refine them. A user study with Ph.D. students indicates that SciCapenter significantly lowers the cognitive load of caption writing. Participants' feedback further offers valuable design insights for future systems aiming to enhance caption writing., Comment: CHI EA '24: Extended Abstracts of the 2024 CHI Conference on Human Factors in Computing Systems

Offline reinforcement learning (RL) presents distinct challenges as it relies solely on observational data. A central concern in this context is ensuring the safety of the learned policy by quantifying uncertainties associated with various actions and environmental stochasticity. Traditional approaches primarily emphasize mitigating epistemic uncertainty by learning risk-averse policies, often overlooking environmental stochasticity. In this study, we propose an uncertainty-aware distributional offline RL method to simultaneously address both epistemic uncertainty and environmental stochasticity. We propose a model-free offline RL algorithm capable of learning risk-averse policies and characterizing the entire distribution of discounted cumulative rewards, as opposed to merely maximizing the expected value of accumulated discounted returns. Our method is rigorously evaluated through comprehensive experiments in both risk-sensitive and risk-neutral benchmarks, demonstrating its superior performance.

Causal inference has shown potential in enhancing the predictive accuracy, fairness, robustness, and explainability of Natural Language Processing (NLP) models by capturing causal relationships among variables. The emergence of generative Large Language Models (LLMs) has significantly impacted various NLP domains, particularly through their advanced reasoning capabilities. This survey focuses on evaluating and improving LLMs from a causal view in the following areas: understanding and improving the LLMs' reasoning capacity, addressing fairness and safety issues in LLMs, complementing LLMs with explanations, and handling multimodality. Meanwhile, LLMs' strong reasoning capacities can in turn contribute to the field of causal inference by aiding causal relationship discovery and causal effect estimations. This review explores the interplay between causal inference frameworks and LLMs from both perspectives, emphasizing their collective potential to further the development of more advanced and equitable artificial intelligence systems.

The long-tail recommendation is a challenging task for traditional recommender systems, due to data sparsity and data imbalance issues. The recent development of large language models (LLMs) has shown their abilities in complex reasoning, which can help to deduce users' preferences based on very few previous interactions. However, since most LLM-based systems rely on items' semantic meaning as the sole evidence for reasoning, the collaborative information of user-item interactions is neglected, which can cause the LLM's reasoning to be misaligned with task-specific collaborative information of the dataset. To further align LLMs' reasoning to task-specific user-item interaction knowledge, we introduce collaborative retrieval-augmented LLMs, CoRAL, which directly incorporate collaborative evidence into the prompts. Based on the retrieved user-item interactions, the LLM can analyze shared and distinct preferences among users, and summarize the patterns indicating which types of users would be attracted by certain items. The retrieved collaborative evidence prompts the LLM to align its reasoning with the user-item interaction patterns in the dataset. However, since the capacity of the input prompt is limited, finding the minimally-sufficient collaborative information for recommendation tasks can be challenging. We propose to find the optimal interaction set through a sequential decision-making process and develop a retrieval policy learned through a reinforcement learning (RL) framework, CoRAL. Our experimental results show that CoRAL can significantly improve LLMs' reasoning abilities on specific recommendation tasks. Our analysis also reveals that CoRAL can more efficiently explore collaborative information through reinforcement learning., Comment: 11 pages

Web-based applications such as chatbots, search engines and news recommendations continue to grow in scale and complexity with the recent surge in the adoption of LLMs. Online model selection has thus garnered increasing attention due to the need to choose the best model among a diverse set while balancing task reward and exploration cost. Organizations faces decisions like whether to employ a costly API-based LLM or a locally finetuned small LLM, weighing cost against performance. Traditional selection methods often evaluate every candidate model before choosing one, which are becoming impractical given the rising costs of training and finetuning LLMs. Moreover, it is undesirable to allocate excessive resources towards exploring poor-performing models. While some recent works leverage online bandit algorithm to manage such exploration-exploitation trade-off in model selection, they tend to overlook the increasing-then-converging trend in model performances as the model is iteratively finetuned, leading to less accurate predictions and suboptimal model selections. In this paper, we propose a time-increasing bandit algorithm TI-UCB, which effectively predicts the increase of model performances due to finetuning and efficiently balances exploration and exploitation in model selection. To further capture the converging points of models, we develop a change detection mechanism by comparing consecutive increase predictions. We theoretically prove that our algorithm achieves a logarithmic regret upper bound in a typical increasing bandit setting, which implies a fast convergence rate. The advantage of our method is also empirically validated through extensive experiments on classification model selection and online selection of LLMs. Our results highlight the importance of utilizing increasing-then-converging pattern for more efficient and economic model selection in the deployment of LLMs., Comment: Accepted by WWW'24 (Oral)

This study explores a minimal renormalizable dark matter (DM) model, incorporating a sub-GeV Majorana DM and a singlet scalar particle $\phi$. Using scalar and pseudo-scalar interactions (couplings $c_s$ and $c_p$), we investigate implications for DM detection, considering $s$-wave, $p$-wave, and combined ($s$+$p$ wave) contributions in DM annihilation cross-section, as well as loop-correction contributions to DM-nucleon elastic scattering. Identifying a broad parameter space ($10 \,\rm{MeV} < m_\chi \lesssim m_\phi$) within the $2\sigma$ allowed region, we explore scenarios ($\left|c_s\right|\gg \left|c_p\right|$, $\left|c_s\right|\ll \left|c_p\right|$, and $\left|c_s\right|\approx \left|c_p\right|$). We find that (i) a non-zero pseudo-scalar coupling alleviates direct detection constraints as a comparison with the previous pure scalar coupling case; (ii) CMB observations set stringent limits on pseudo-scalar interaction dominant cases, making $s$-wave annihilation viable only for $m_\chi>1\,\rm{GeV}$; (iii) the preferred $\phi$-resonance region can be tested in the future indirect detection experiments, such as e-ASTROGAM., Comment: 35 pages, 4 figures

Large Language Models (LLMs) have been widely used as general-purpose AI agents showing comparable performance on many downstream tasks. However, existing work shows that it is challenging for LLMs to integrate structured data (e.g. KG, tables, DBs) into their prompts; LLMs need to either understand long text data or select the most relevant evidence prior to inference, and both approaches are not trivial. In this paper, we propose a framework, Learning to Reduce, that fine-tunes a language model to generate a reduced version of an input context, given a task description and context input. The model learns to reduce the input context using On-Policy Reinforcement Learning and aims to improve the reasoning performance of a fixed LLM. Experimental results illustrate that our model not only achieves comparable accuracies in selecting the relevant evidence from an input context, but also shows generalizability on different datasets. We further show that our model helps improve the LLM's performance on downstream tasks especially when the context is long., Comment: 5 pages

Recently, Foundation Models (FMs), with their extensive knowledge bases and complex architectures, have offered unique opportunities within the realm of recommender systems (RSs). In this paper, we attempt to thoroughly examine FM-based recommendation systems (FM4RecSys). We start by reviewing the research background of FM4RecSys. Then, we provide a systematic taxonomy of existing FM4RecSys research works, which can be divided into four different parts including data characteristics, representation learning, model type, and downstream tasks. Within each part, we review the key recent research developments, outlining the representative models and discussing their characteristics. Moreover, we elaborate on the open problems and opportunities of FM4RecSys aiming to shed light on future research directions in this area. In conclusion, we recap our findings and discuss the emerging trends in this field.

Large language models (LLMs) have shown remarkable advances in language generation and understanding but are also prone to exhibiting harmful social biases. While recognition of these behaviors has generated an abundance of bias mitigation techniques, most require modifications to the training data, model parameters, or decoding strategy, which may be infeasible without access to a trainable model. In this work, we leverage the zero-shot capabilities of LLMs to reduce stereotyping in a technique we introduce as zero-shot self-debiasing. With two approaches, self-debiasing via explanation and self-debiasing via reprompting, we show that self-debiasing can significantly reduce the degree of stereotyping across nine different social groups while relying only on the LLM itself and a simple prompt, with explanations correctly identifying invalid assumptions and reprompting delivering the greatest reductions in bias. We hope this work opens inquiry into other zero-shot techniques for bias mitigation.

Table question answering is a popular task that assesses a model's ability to understand and interact with structured data. However, the given table often does not contain sufficient information for answering the question, necessitating the integration of external knowledge. Existing methods either convert both the table and external knowledge into text, which neglects the structured nature of the table; or they embed queries for external sources in the interaction with the table, which complicates the process. In this paper, we propose a simple yet effective method to integrate external information in a given table. Our method first constructs an augmenting table containing the missing information and then generates a SQL query over the two tables to answer the question. Experiments show that our method outperforms strong baselines on three table QA benchmarks. Our code is publicly available at https://github.com/UCSB-NLP-Chang/Augment_tableQA.

Large pretrained multilingual language models (ML-LMs) have shown remarkable capabilities of zero-shot cross-lingual transfer, without direct cross-lingual supervision. While these results are promising, follow-up works found that, within the multilingual embedding spaces, there exists strong language identity information which hinders the expression of linguistic factors shared across languages. For semantic tasks like cross-lingual sentence retrieval, it is desired to remove such language identity signals to fully leverage semantic information. In this work, we provide a novel view of projecting away language-specific factors from a multilingual embedding space. Specifically, we discover that there exists a low-rank subspace that primarily encodes information irrelevant to semantics (e.g., syntactic information). To identify this subspace, we present a simple but effective unsupervised method based on singular value decomposition with multiple monolingual corpora as input. Once the subspace is found, we can directly project the original embeddings into the null space to boost language agnosticism without finetuning. We systematically evaluate our method on various tasks including the challenging language-agnostic QA retrieval task. Empirical results show that applying our method consistently leads to improvements over commonly used ML-LMs., Comment: 17 pages, 7 figures, EMNLP 2022 (main conference)

Stable transport of laser beams in highly over-dense plasmas is of significance in fast ignition of inertial confinement fusion, relativistic electrons generation, and powerful electromagnetic emission, but hard to realize. Early in 1996, Harris proposed an electromagnetically induced transparency (EIT) mechanism, analogous to the concept in atom physics, to transport a low-frequency (LF) laser in over-dense plasmas aided by a high-frequency pump laser. However, subsequent investigations show that EIT cannot occur in real plasmas with boundaries. Here, our particle-in-cell simulations show that EIT can occur in strongly-relativistic regime and results in stable propagation of a LF laser in bounded plasmas with tens of its critical density. A relativistic three-wave coupling model is developed, and the criteria and frequency passband for EIT occurrence are presented. The passband is sufficiently wide in strongly-relativistic regime, allowing EIT to work sustainably. Nevertheless, it is narrowed to nearly an isolated point in weakly-relativistic regime, which can explain the quenching of EIT in bounded plasmas found in previous investigations.

The recognition of underwater audio plays a significant role in identifying a vessel while it is in motion. Underwater target recognition tasks have a wide range of applications in areas such as marine environmental protection, detection of ship radiated noise, underwater noise control, and coastal vessel dispatch. The traditional UATR task involves training a network to extract features from audio data and predict the vessel type. The current UATR dataset exhibits shortcomings in both duration and sample quantity. In this paper, we propose Oceanship, a large-scale and diverse underwater audio dataset. This dataset comprises 15 categories, spans a total duration of 121 hours, and includes comprehensive annotation information such as coordinates, velocity, vessel types, and timestamps. We compiled the dataset by crawling and organizing original communication data from the Ocean Communication Network (ONC) database between 2021 and 2022. While audio retrieval tasks are well-established in general audio classification, they have not been explored in the context of underwater audio recognition. Leveraging the Oceanship dataset, we introduce a baseline model named Oceannet for underwater audio retrieval. This model achieves a recall at 1 (R@1) accuracy of 67.11% and a recall at 5 (R@5) accuracy of 99.13% on the Deepship dataset., Comment: Accepted by ICIC 2024

In this work, we develop and release Yuan 2.0, a series of large language models with parameters ranging from 2.1 billion to 102.6 billion. The Localized Filtering-based Attention (LFA) is introduced to incorporate prior knowledge of local dependencies of natural language into Attention. A data filtering and generating system is presented to build pre-training and fine-tuning dataset in high quality. A distributed training method with non-uniform pipeline parallel, data parallel, and optimizer parallel is proposed, which greatly reduces the bandwidth requirements of intra-node communication, and achieves good performance in large-scale distributed training. Yuan 2.0 models display impressive ability in code generation, math problem-solving, and chatting compared with existing models. The latest version of YUAN 2.0, including model weights and source code, is accessible at Github.

After the successful fusion ignition at National Ignition Facility, seeking for a high-gain fusion scheme becomes the next hot-spot in inertial confinement fusion community. Fast ignition provides an alternative due to its potential to reduce the energy of driven lasers and achieve higher target gain, whose core step is generating fast electron beam using picosecond lasers. The properties of the electron beam, such as energy spectra, determines the succeed heating process and the yield. In this study, with 2D particle-in-cell simulations, we find that the energy transfer ratio would significantly increase, and the energy spectra would be lifted up in the dual-color injecting scheme, where an extra weak low-frequency laser is injected with the main pulse. These phenomena are attributed to the longitudinal electric field modulation in low-density region and electromagnetically-induced-transparency process in the microchannels formed near the critical surface. Our results can be applied to fast ignition schemes and experimental designs of super-hot electron generation with different pulse durations. For example, in the DCI project with a fast ignition scheme, such a dual-color intense picosecond laser facility is under construction to control the fast electron generation and enhance the laser transparency., Comment: 24 pages, 8 figures

Relativistic positron sources with high spin polarization have important applications in nuclear and particle physics and many frontier fields. However, it is challenging to produce dense polarized positrons. Here we present a simple and effective method to achieve such a positron source by directly impinging a relativistic high-density electron beam on the surface of a solid target. During the interaction, a strong return current of plasma electrons is induced and subsequently asymmetric quasistatic magnetic fields as high as megatesla are generated along the target surface. This gives rise to strong radiative spin flips and multiphoton processes, thus leading to efficient generation of copious polarized positrons. With three-dimensional particle-in-cell simulations, we demonstrate the production of a dense highly-polarized multi-GeV positron beam with an average spin polarization above 40% and nC-scale charge per shot. This offers a novel route for the studies of laserless strong-field quantum electrodynamics physics and for the development of high-energy polarized positron sources.

In kernel-centric operations, the uprobe component of eBPF frequently encounters performance bottlenecks, largely attributed to the overheads borne by context switches. Transitioning eBPF operations to user space bypasses these hindrances, thereby optimizing performance. This also enhances configurability and obviates the necessity for root access or privileges for kernel eBPF, subsequently minimizing the kernel attack surface. This paper introduces bpftime, a novel user-space eBPF runtime, which leverages binary rewriting to implement uprobe and syscall hook capabilities. Through bpftime, userspace uprobes achieve a 10x speed enhancement compared to their kernel counterparts without requiring dual context switches. Additionally, this runtime facilitates the programmatic hooking of syscalls within a process, both safely and efficiently. Bpftime can be seamlessly attached to any running process, limiting the need for either a restart or manual recompilation. Our implementation also extends to interprocess eBPF Maps within shared memory, catering to summary aggregation or control plane communication requirements. Compatibility with existing eBPF toolchains such as clang and libbpf is maintained, not only simplifying the development of user-space eBPF without necessitating any modifications but also supporting CO-RE through BTF. Through bpftime, we not only enhance uprobe performance but also extend the versatility and user-friendliness of eBPF runtime in user space, paving the way for more efficient and secure kernel operations.