Search

While large language models (LLMs) present significant potential for supporting numerous real-world applications and delivering positive social impacts, they still face significant challenges in terms of the inherent risk of privacy leakage, hallucinated outputs, and value misalignment, and can be maliciously used for generating toxic content and unethical purposes after been jailbroken. Therefore, in this survey, we present a comprehensive review of recent advancements aimed at mitigating these issues, organized across the four phases of LLM development and usage: data collecting and pre-training, fine-tuning and alignment, prompting and reasoning, and post-processing and auditing. We elaborate on the recent advances for enhancing the performance of LLMs in terms of privacy protection, hallucination reduction, value alignment, toxicity elimination, and jailbreak defenses. In contrast to previous surveys that focus on a single dimension of responsible LLMs, this survey presents a unified framework that encompasses these diverse dimensions, providing a comprehensive view of enhancing LLMs to better serve real-world applications.

This study investigates the hydrodynamic drag force on side-by-side flexible blades in a uniform steady current through experimental and theoretical approaches. Four different blade mimics were arranged in side-by-side bunches and tested in a circulating water tunnel. The experiments cover a static regime and a dynamic instability regime known as flutter. We examine four non-dimensional parameters to assess their effects on the bulk drag coefficient $C_{D,\mathrm{bulk}}$ and the onset of flutter: the drag-to-stiffness ratio $\mathrm{Ca}$, the buoyancy-to-stiffness ratio $\mathrm{B}$, the mass ratio of fluid inertia to total system inertia $\beta$, and the slenderness parameter $\lambda$, which represents the ratio of the resistive drag to the reactive force. The results show that $C_{D,\mathrm{bulk}}$ decreases in the static regime starting at $\mathrm{Ca}/B > \textit{O}(1)$ and settles to an almost constant value in the flutter regime at high $\mathrm{Ca}$. In the static regime, $\mathrm{B}$ is the primary influencing factor. Increasing $\beta$, $\mathrm{B}$, or $\lambda$ stabilizes the system and delays the onset of flutter. By introducing an equivalent thickness and bending stiffness for a bunch of blades, we utilize well-established analytical and numerical models for individual blades to predict the drag reduction of side-by-side blade assemblies. The analytical model accurately predicts drag reduction in the static regime, while the numerical model effectively predicts both the onset of flutter and drag reduction across both regimes with appropriate cross-flow hydrodynamic coefficients. Meanwhile, we investigate the reactive force terms to unveil their impact on the system stability and drag reduction, demonstrating its superiority over the traditional Morison's equation for highly compliant blades in cross-flow scenarios.

For the particle detectors based on noble gases or liquids, it is essential to understand the transport dynamic and the properties of the electrons. We report the development of a tool for electron transport in noble gases He, Ne, Ar, Kr, or Xe, and liquids Ar, Kr, or Xe. The simulation, implemented in C++ and MATLAB, is based on electron-atom collisions, including elastic scattering, excitation and ionization. We validate the program through assessing the electron's swarm parameters, specifically the drift velocity and the diffusion coefficient. For electron transport in liquids, two models are discussed and both are used for the construction of the Monte Carlo framework based on the Cohen Leker theory. The results demonstrate the effectiveness and accuracy of the simulation tool, which offers a valuable support for detector design and data analysis.

Video scene detection involves assessing whether each shot and its surroundings belong to the same scene. Achieving this requires meticulously correlating multi-modal cues, $\it{e.g.}$ visual entity and place modalities, among shots and comparing semantic changes around each shot. However, most methods treat multi-modal semantics equally and do not examine contextual differences between the two sides of a shot, leading to sub-optimal detection performance. In this paper, we propose the $\bf{M}$odality-$\bf{A}$ware $\bf{S}$hot $\bf{R}$elating and $\bf{C}$omparing approach (MASRC), which enables relating shots per their own characteristics of visual entity and place modalities, as well as comparing multi-shots similarities to have scene changes explicitly encoded. Specifically, to fully harness the potential of visual entity and place modalities in modeling shot relations, we mine long-term shot correlations from entity semantics while simultaneously revealing short-term shot correlations from place semantics. In this way, we can learn distinctive shot features that consolidate coherence within scenes and amplify distinguishability across scenes. Once equipped with distinctive shot features, we further encode the relations between preceding and succeeding shots of each target shot by similarity convolution, aiding in the identification of scene ending shots. We validate the broad applicability of the proposed components in MASRC. Extensive experimental results on public benchmark datasets demonstrate that the proposed MASRC significantly advances video scene detection.

Retrieval-augmented generation (RAG) synergizes the retrieval of pertinent data with the generative capabilities of Large Language Models (LLMs), ensuring that the generated output is not only contextually relevant but also accurate and current. We introduce XRAG, an open-source, modular codebase that facilitates exhaustive evaluation of the performance of foundational components of advanced RAG modules. These components are systematically categorized into four core phases: pre-retrieval, retrieval, post-retrieval, and generation. We systematically analyse them across reconfigured datasets, providing a comprehensive benchmark for their effectiveness. As the complexity of RAG systems continues to escalate, we underscore the critical need to identify potential failure points in RAG systems. We formulate a suite of experimental methodologies and diagnostic testing protocols to dissect the failure points inherent in RAG engineering. Subsequently, we proffer bespoke solutions aimed at bolstering the overall performance of these modules. Our work thoroughly evaluates the performance of advanced core components in RAG systems, providing insights into optimizations for prevalent failure points.

Connected component (CC) is a proper text shape representation that aligns with human reading intuition. However, CC-based text detection methods have recently faced a developmental bottleneck that their time-consuming post-processing is difficult to eliminate. To address this issue, we introduce an explicit relational reasoning network (ERRNet) to elegantly model the component relationships without post-processing. Concretely, we first represent each text instance as multiple ordered text components, and then treat these components as objects in sequential movement. In this way, scene text detection can be innovatively viewed as a tracking problem. From this perspective, we design an end-to-end tracking decoder to achieve a CC-based method dispensing with post-processing entirely. Additionally, we observe that there is an inconsistency between classification confidence and localization quality, so we propose a Polygon Monte-Carlo method to quickly and accurately evaluate the localization quality. Based on this, we introduce a position-supervised classification loss to guide the task-aligned learning of ERRNet. Experiments on challenging benchmarks demonstrate the effectiveness of our ERRNet. It consistently achieves state-of-the-art accuracy while holding highly competitive inference speed., Comment: Accepted to AAAI 2025

This paper proposes a dual-color grating chip design method for simultaneously capturing dual atomic clouds (87Rb and 133Cs). By simulating key parameters such as the grating period, etching depth, duty cycle, coating material, and thickness, the optimal design parameters were determined to ensure efficient dual-wavelength diffraction and maximize the number of captured atoms. Experimental results demonstrate the simultaneous trapping of 1.6E8 87Rb atoms and 7.8E6 133Cs atoms, thereby offering an approach for multi-species cold atom systems. This dual-species grating magneto-optical trap (GMOT) system has potential applications in precision measurements such as cold atom clocks, quantum interferometers, and quantum electrometry.

Physics problems constitute a significant aspect of reasoning, necessitating complicated reasoning ability and abundant physics knowledge. However, existing large language models (LLMs) frequently fail due to a lack of knowledge or incorrect knowledge application. To mitigate these issues, we propose Physics Reasoner, a knowledge-augmented framework to solve physics problems with LLMs. Specifically, the proposed framework constructs a comprehensive formula set to provide explicit physics knowledge and utilizes checklists containing detailed instructions to guide effective knowledge application. Namely, given a physics problem, Physics Reasoner solves it through three stages: problem analysis, formula retrieval, and guided reasoning. During the process, checklists are employed to enhance LLMs' self-improvement in the analysis and reasoning stages. Empirically, Physics Reasoner mitigates the issues of insufficient knowledge and incorrect application, achieving state-of-the-art performance on SciBench with an average accuracy improvement of 5.8%., Comment: COLING 2025

Machine learning training has emerged as one of the most prominent workloads in modern data centers. These training jobs are large-scale, long-lasting, and tightly coupled, and are often disrupted by various events in the cluster such as failures, maintenance, and job scheduling. To handle these events, we rely on cold migration, where we first checkpoint the entire cluster, replace the related machines, and then restart the training. This approach leads to disruptions to the training jobs, resulting in significant downtime. In this paper, we present TrainMover, a live migration system that enables machine replacement during machine learning training. TrainMover minimizes downtime by leveraging member replacement of collective communication groups and sandbox lazy initialization. Our evaluation demonstrates that TrainMover achieves 16x less downtime compared to all baselines, effectively handling data center events like straggler rebalancing, maintenance, and unexpected failures., Comment: 13 pages body, 19 pages total

Simulating the nature of quantum fields in diverse spacetime backgrounds offers valuable insights for the fundamental comprehension of quantum mechanics and general relativity. Here we introduce a novel method for mapping the massless Dirac equation in 1+1D curved spacetime to a controllable quantum simulation model, applicable to various observers' perspectives. We perform numerical simulations of Simpson spacetime and calculate tunneling rates in Painleve and Schwarzschild coordinates, which align closely with theoretical predictions of Hawking radiation. Additionally, we show the transition of Simpson spacetime from a regular black hole to a wormhole as the parameter $"a > r_s"$. This method facilitates the study of spacetime from various coordinate perspectives (observers), providing deeper insights and understanding., Comment: 7 pages, 10 figures

Structure-based molecule optimization (SBMO) aims to optimize molecules with both continuous coordinates and discrete types against protein targets. A promising direction is to exert gradient guidance on generative models given its remarkable success in images, but it is challenging to guide discrete data and risks inconsistencies between modalities. To this end, we leverage a continuous and differentiable space derived through Bayesian inference, presenting Molecule Joint Optimization (MolJO), the first gradient-based SBMO framework that facilitates joint guidance signals across different modalities while preserving SE(3)-equivariance. We introduce a novel backward correction strategy that optimizes within a sliding window of the past histories, allowing for a seamless trade-off between explore-and-exploit during optimization. Our proposed MolJO achieves state-of-the-art performance on CrossDocked2020 benchmark (Success Rate 51.3% , Vina Dock -9.05 and SA 0.78), more than 4x improvement in Success Rate compared to the gradient-based counterpart, and 2x "Me-Better" Ratio as much as 3D baselines. Furthermore, we extend MolJO to a wide range of optimization settings, including multi-objective optimization and challenging tasks in drug design such as R-group optimization and scaffold hopping, further underscoring its versatility and potential., Comment: 27 pages, 17 figures

Given the inherent non-stationarity prevalent in real-world applications, continual Reinforcement Learning (RL) aims to equip the agent with the capability to address a series of sequentially presented decision-making tasks. Within this problem setting, a pivotal challenge revolves around \textit{catastrophic forgetting} issue, wherein the agent is prone to effortlessly erode the decisional knowledge associated with past encountered tasks when learning the new one. In recent progresses, the \textit{generative replay} methods have showcased substantial potential by employing generative models to replay data distribution of past tasks. Compared to storing the data from past tasks directly, this category of methods circumvents the growing storage overhead and possible data privacy concerns. However, constrained by the expressive capacity of generative models, existing \textit{generative replay} methods face challenges in faithfully reconstructing the data distribution of past tasks, particularly in scenarios with a myriad of tasks or high-dimensional data. Inspired by the success of diffusion models in various generative tasks, this paper introduces a novel continual RL algorithm DISTR (Diffusion-based Trajectory Replay) that employs a diffusion model to memorize the high-return trajectory distribution of each encountered task and wakeups these distributions during the policy learning on new tasks. Besides, considering the impracticality of replaying all past data each time, a prioritization mechanism is proposed to prioritize the trajectory replay of pivotal tasks in our method. Empirical experiments on the popular continual RL benchmark \texttt{Continual World} demonstrate that our proposed method obtains a favorable balance between \textit{stability} and \textit{plasticity}, surpassing various existing continual RL baselines in average success rate., Comment: 10 pages, 3 figures, 1 table, inclusion at ICLR 2024 Workshop on Generative Models for Decision Making

World models play a crucial role in decision-making within embodied environments, enabling cost-free explorations that would otherwise be expensive in the real world. To facilitate effective decision-making, world models must be equipped with strong generalizability to support faithful imagination in out-of-distribution (OOD) regions and provide reliable uncertainty estimation to assess the credibility of the simulated experiences, both of which present significant challenges for prior scalable approaches. This paper introduces WHALE, a framework for learning generalizable world models, consisting of two key techniques: behavior-conditioning and retracing-rollout. Behavior-conditioning addresses the policy distribution shift, one of the primary sources of the world model generalization error, while retracing-rollout enables efficient uncertainty estimation without the necessity of model ensembles. These techniques are universal and can be combined with any neural network architecture for world model learning. Incorporating these two techniques, we present Whale-ST, a scalable spatial-temporal transformer-based world model with enhanced generalizability. We demonstrate the superiority of Whale-ST in simulation tasks by evaluating both value estimation accuracy and video generation fidelity. Additionally, we examine the effectiveness of our uncertainty estimation technique, which enhances model-based policy optimization in fully offline scenarios. Furthermore, we propose Whale-X, a 414M parameter world model trained on 970K trajectories from Open X-Embodiment datasets. We show that Whale-X exhibits promising scalability and strong generalizability in real-world manipulation scenarios using minimal demonstrations.

As a prominent category of imitation learning methods, adversarial imitation learning (AIL) has garnered significant practical success powered by neural network approximation. However, existing theoretical studies on AIL are primarily limited to simplified scenarios such as tabular and linear function approximation and involve complex algorithmic designs that hinder practical implementation, highlighting a gap between theory and practice. In this paper, we explore the theoretical underpinnings of online AIL with general function approximation. We introduce a new method called optimization-based AIL (OPT-AIL), which centers on performing online optimization for reward functions and optimism-regularized Bellman error minimization for Q-value functions. Theoretically, we prove that OPT-AIL achieves polynomial expert sample complexity and interaction complexity for learning near-expert policies. To our best knowledge, OPT-AIL is the first provably efficient AIL method with general function approximation. Practically, OPT-AIL only requires the approximate optimization of two objectives, thereby facilitating practical implementation. Empirical studies demonstrate that OPT-AIL outperforms previous state-of-the-art deep AIL methods in several challenging tasks., Comment: Published in NeurIPS 2024: Tian Xu, Zhilong Zhang, Ruishuo Chen, Yihao Sun, Yang Yu. Provably and practically efficient adversarial imitation learning with general function approximation. In: Advances in Neural Information Processing Systems 38 (NeurIPS'24), Vancouver, Canada, 2024

The spontaneous conversion of muonium to antimuonium is one of the interesting charged lepton flavor violation phenomena, offering a sensitive probe of potential new physics and serving as a tool to constrain the parameter space beyond the Standard Model. Utilizing a high-intensity muon beam, a Michel electron magnetic spectrometer and a positron transport solenoid together with a positron detection system, MACE aims to discover or constrain this rare process at the conversion probability beyond the level of $10^{-13}$. This report provides an overview of the theoretical framework and detailed experimental design in the search for the muonium-to-antimuonium conversion., Comment: 115 pages, 59 figures

Wavefield reconstruction inversion (WRI) has been considered a potential solution to the issue of local minima inherent in conventional full waveform inversion (FWI) methods. However, most current WRI research has been confined to 2D problems due to the computational challenges posed by solving augmented systems for optimal data-fitting wavefields. This constraint limits WRI applicability to realistic 3D scenarios. This study introduces a GPU-accelerated 3D source-free adaptive WRI (GPU-SF-AWRI) method that overcomes these computational barriers by adaptively controlling the computational accuracy of wavefield simulation and optimizing GPU utilization, thus enhancing its suitability for 3D applications. The inclusion of an on-the-fly source estimation technique further boosts its performance on realistic problems. Numerical experiments reveal that the proposed GPU-accelerated method achieves a 195-fold speedup compared to CPU-based approaches. By incorporating adaptive accuracy and total variation regularization, we attain a 2-fold speedup while maintaining inversion accuracy. We applied the GPU-SF-AWRI method to numerical and actual Vertical Seismic Profiling (VSP) physical modeling P-wave data, confirming its efficacy in addressing real data challenges and mitigating local minima associated with conventional FWI.

In this paper, we investigate a class of inviscid generalized surface quasi-geostrophic (SQG) equations on the half-plane with a rigid boundary. Compared to the Biot-Savart law in the vorticity form of the 2D Euler equation, the velocity formula here includes an additional Fourier multiplier operator $m(\Lambda)$. When $m(\Lambda) = \Lambda^\alpha$, where $\Lambda = (-\Delta)^{1/2}$ and $\alpha\in (0,2)$, the equation reduces to the well-known $\alpha$-SQG equation. Finite-time singularity formation for patch solutions to the $\alpha$-SQG equation was famously discovered by Kiselev, Ryzhik, Yao, and Zlato\v{s} [Ann. Math., 184 (2016), pp. 909-948]. We establish finite-time singularity formation for patch solutions to the generalized SQG equations under the Osgood condition \[\int_2^\infty \frac{1}{r (\log r) m(r)} dr < \infty\] along with some additional mild conditions. Notably, our result fills the gap between the globally well-posed 2D Euler equation ($\alpha = 0$) and the $\alpha$-SQG equation ($\alpha > 0$). Furthermore, in line with Elgindi's global regularity results for 2D Loglog-Euler type equations [Arch. Rat. Mech. Anal., 211 (2014), pp. 965-990], our findings suggest that the Osgood condition serves as a sharp threshold that distinguishes global regularity and finite-time singularity in these models. In addition, we generalize the local regularity and finite-time singularity results for patch solutions to the $\alpha$-SQG equation, as established by Gancedo and Patel [Ann. PDE, 7 (2021), no. 1, Art. no. 4], extending them to cases where $m(r)$ behaves like $r^\alpha$ near infinity but does not have an explicit formulation., Comment: 62 pages, 4 figures

The emerging short-form video platforms have been growing tremendously and become one of the leading social media recently. Although the expanded popularity of these platforms has attracted increasing research attention, there has been a lack of understanding of whether and how they deviate from traditional long-form video-sharing platforms such as YouTube and Bilibili. To address this, we conduct a large-scale data-driven analysis of Kuaishou, one of the largest short-form video platforms in China. Based on 248 million videos uploaded to the platform across all categories, we identify their notable differences from long-form video platforms through a comparison study with Bilibili, a leading long-form video platform in China. We find that videos are shortened by multiples on Kuaishou, with distinctive categorical distributions over-represented by life-related rather than interest-based videos. Users interact with videos less per view, but top videos can even more effectively acquire users' collective attention. More importantly, ordinary content creators have higher probabilities of producing hit videos. Our results shed light on the uniqueness of short-form video platforms and pave the way for future research and design for better short-form video ecology.

There is growing research interest in measuring the statistical heterogeneity of clients' local datasets. Such measurements are used to estimate the suitability for collaborative training of personalized federated learning (PFL) models. Currently, these research endeavors are taking place in silos and there is a lack of a unified benchmark to provide a fair and convenient comparison among various approaches in common settings. We aim to bridge this important gap in this paper. The proposed benchmarking framework currently includes six representative approaches. Extensive experiments have been conducted to compare these approaches under five standard non-IID FL settings, providing much needed insights into which approaches are advantageous under which settings. The proposed framework offers useful guidance on the suitability of various data divergence measures in FL systems. It is beneficial for keeping related research activities on the right track in terms of: (1) designing PFL schemes, (2) selecting appropriate data heterogeneity evaluation approaches for specific FL application scenarios, and (3) addressing fairness issues in collaborative model training. The code is available at https://github.com/Xiaoni-61/DH-Benchmark., Comment: Accepted to FL@FM-NeurIPS'24

Artistic typography is a technique to visualize the meaning of input character in an imaginable and readable manner. With powerful text-to-image diffusion models, existing methods directly design the overall geometry and texture of input character, making it challenging to ensure both creativity and legibility. In this paper, we introduce a dual-branch and training-free method, namely VitaGlyph, enabling flexible artistic typography along with controllable geometry change to maintain the readability. The key insight of VitaGlyph is to treat input character as a scene composed of Subject and Surrounding, followed by rendering them under varying degrees of geometry transformation. The subject flexibly expresses the essential concept of input character, while the surrounding enriches relevant background without altering the shape. Specifically, we implement VitaGlyph through a three-phase framework: (i) Knowledge Acquisition leverages large language models to design text descriptions of subject and surrounding. (ii) Regional decomposition detects the part that most matches the subject description and divides input glyph image into subject and surrounding regions. (iii) Typography Stylization firstly refines the structure of subject region via Semantic Typography, and then separately renders the textures of Subject and Surrounding regions through Controllable Compositional Generation. Experimental results demonstrate that VitaGlyph not only achieves better artistry and readability, but also manages to depict multiple customize concepts, facilitating more creative and pleasing artistic typography generation. Our code will be made publicly at https://github.com/Carlofkl/VitaGlyph., Comment: https://github.com/Carlofkl/VitaGlyph

Optical imaging systems are generally limited by the depth of field because of the nature of the optics. Therefore, extending depth of field (EDoF) is a fundamental task for meeting the requirements of emerging visual applications. To solve this task, the common practice is using multi-focus images from a single viewpoint. This method can obtain acceptable quality of EDoF under the condition of fixed field of view, but it is only applicable to static scenes and the field of view is limited and fixed. An emerging data type, varifocal multiview images have the potential to become a new paradigm for solving the EDoF, because the data contains more field of view information than multi-focus images. To realize EDoF of varifocal multiview images, we propose an end-to-end method for the EDoF, including image alignment, image optimization and image fusion. Experimental results demonstrate the efficiency of the proposed method.

The current high-dimensional linear factor models fail to account for the different types of variables, while high-dimensional nonlinear factor models often overlook the overdispersion present in mixed-type data. However, overdispersion is prevalent in practical applications, particularly in fields like biomedical and genomics studies. To address this practical demand, we propose an overdispersed generalized factor model (OverGFM) for performing high-dimensional nonlinear factor analysis on overdispersed mixed-type data. Our approach incorporates an additional error term to capture the overdispersion that cannot be accounted for by factors alone. However, this introduces significant computational challenges due to the involvement of two high-dimensional latent random matrices in the nonlinear model. To overcome these challenges, we propose a novel variational EM algorithm that integrates Laplace and Taylor approximations. This algorithm provides iterative explicit solutions for the complex variational parameters and is proven to possess excellent convergence properties. We also develop a criterion based on the singular value ratio to determine the optimal number of factors. Numerical results demonstrate the effectiveness of this criterion. Through comprehensive simulation studies, we show that OverGFM outperforms state-of-the-art methods in terms of estimation accuracy and computational efficiency. Furthermore, we demonstrate the practical merit of our method through its application to two datasets from genomics. To facilitate its usage, we have integrated the implementation of OverGFM into the R package GFM.

Jailbreak attacks on Language Model Models (LLMs) entail crafting prompts aimed at exploiting the models to generate malicious content. This paper proposes a new type of jailbreak attacks which shift the attention of the LLM by inserting a prohibited query into a carrier article. The proposed attack leverage the knowledge graph and a composer LLM to automatically generating a carrier article that is similar to the topic of the prohibited query but does not violate LLM's safeguards. By inserting the malicious query to the carrier article, the assembled attack payload can successfully jailbreak LLM. To evaluate the effectiveness of our method, we leverage 4 popular categories of ``harmful behaviors'' adopted by related researches to attack 6 popular LLMs. Our experiment results show that the proposed attacking method can successfully jailbreak all the target LLMs which high success rate, except for Claude-3.

Time Series Classification (TSC) encompasses two settings: classifying entire sequences or classifying segmented subsequences. The raw time series for segmented TSC usually contain Multiple classes with Varying Duration of each class (MVD). Therefore, the characteristics of MVD pose unique challenges for segmented TSC, yet have been largely overlooked by existing works. Specifically, there exists a natural temporal dependency between consecutive instances (segments) to be classified within MVD. However, mainstream TSC models rely on the assumption of independent and identically distributed (i.i.d.), focusing on independently modeling each segment. Additionally, annotators with varying expertise may provide inconsistent boundary labels, leading to unstable performance of noise-free TSC models. To address these challenges, we first formally demonstrate that valuable contextual information enhances the discriminative power of classification instances. Leveraging the contextual priors of MVD at both the data and label levels, we propose a novel consistency learning framework Con4m, which effectively utilizes contextual information more conducive to discriminating consecutive segments in segmented TSC tasks, while harmonizing inconsistent boundary labels for training. Extensive experiments across multiple datasets validate the effectiveness of Con4m in handling segmented TSC tasks on MVD.

Combining offline and online reinforcement learning (RL) techniques is indeed crucial for achieving efficient and safe learning where data acquisition is expensive. Existing methods replay offline data directly in the online phase, resulting in a significant challenge of data distribution shift and subsequently causing inefficiency in online fine-tuning. To address this issue, we introduce an innovative approach, \textbf{E}nergy-guided \textbf{DI}ffusion \textbf{S}ampling (EDIS), which utilizes a diffusion model to extract prior knowledge from the offline dataset and employs energy functions to distill this knowledge for enhanced data generation in the online phase. The theoretical analysis demonstrates that EDIS exhibits reduced suboptimality compared to solely utilizing online data or directly reusing offline data. EDIS is a plug-in approach and can be combined with existing methods in offline-to-online RL setting. By implementing EDIS to off-the-shelf methods Cal-QL and IQL, we observe a notable 20% average improvement in empirical performance on MuJoCo, AntMaze, and Adroit environments. Code is available at \url{https://github.com/liuxhym/EDIS}.

Accessing the synthesizability of crystal structures is pivotal for advancing the practical application of theoretical material structures designed by machine learning or high-throughput screening. However, a significant gap exists between the actual synthesizability and thermodynamic or kinetic stability, which is commonly used for screening theoretical structures for experiments. To address this, we develop the Crystal Synthesis Large Language Models (CSLLM) framework, which includes three LLMs for predicting the synthesizability, synthesis methods, and precursors. We create a comprehensive synthesizability dataset including 140,120 crystal structures and develop an efficient text representation method for crystal structures to fine-tune the LLMs. The Synthesizability LLM achieves a remarkable 98.6% accuracy, significantly outperforming traditional synthesizability screening based on thermodynamic and kinetic stability by 106.1% and 44.5%, respectively. The Methods LLM achieves a classification accuracy of 91.02%, and the Precursors LLM has an 80.2% success rate in predicting synthesis precursors. Furthermore, we develop a user-friendly graphical interface that enables automatic predictions of synthesizability and precursors from uploaded crystal structure files. Through these contributions, CSLLM bridges the gap between theoretical material design and experimental synthesis, paving the way for the rapid discovery of novel and synthesizable functional materials.

Artificial Intelligence-Generated Content (AIGC)-content autonomously produced by AI systems without human intervention-has significantly boosted efficiency across various fields. However, the AIGC in material science faces challenges in the ability to efficiently discover innovative materials that surpass existing databases, alongside the invariances and stability considerations of crystal structures. To address these challenges, we develop T2MAT (Text-to-Material), a comprehensive framework processing from a user-input sentence to inverse design material structures with goal properties beyond the existing database via globally exploring chemical space, followed by an entirely automated workflow of first principal validation. Furthermore, we propose CGTNet (Crystal Graph Transformer NETwork), a novel graph neural network model that captures long-term interactions, to enhance the accuracy and data efficiency of property prediction and thereby improve the reliability of inverse design. Through these contributions, T2MAT minimizes the dependency on human expertise and significantly enhances the efficiency of designing novel, high-performance functional materials, thereby actualizing AIGC in the materials design domain.

Flow matching (FM) is a general framework for defining probability paths via Ordinary Differential Equations (ODEs) to transform between noise and data samples. Recent approaches attempt to straighten these flow trajectories to generate high-quality samples with fewer function evaluations, typically through iterative rectification methods or optimal transport solutions. In this paper, we introduce Consistency Flow Matching (Consistency-FM), a novel FM method that explicitly enforces self-consistency in the velocity field. Consistency-FM directly defines straight flows starting from different times to the same endpoint, imposing constraints on their velocity values. Additionally, we propose a multi-segment training approach for Consistency-FM to enhance expressiveness, achieving a better trade-off between sampling quality and speed. Preliminary experiments demonstrate that our Consistency-FM significantly improves training efficiency by converging 4.4x faster than consistency models and 1.7x faster than rectified flow models while achieving better generation quality. Our code is available at: https://github.com/YangLing0818/consistency_flow_matching, Comment: Code: https://github.com/YangLing0818/consistency_flow_matching

In this paper, we investigate the amalgamation of cell-free (CF) and extremely large-scale multiple-input multiple-output (XL-MIMO) technologies, referred to as a CF XL-MIMO, as a promising advancement for enabling future mobile networks. To address the computational complexity and communication power consumption associated with conventional centralized optimization, we focus on user-centric dynamic networks in which each user is served by an adaptive subset of access points (AP) rather than all of them. We begin our research by analyzing a joint resource allocation problem for energy-efficient CF XL-MIMO systems, encompassing cooperative clustering and power control design, where all clusters are adaptively adjustable. Then, we propose an innovative double-layer multi-agent reinforcement learning (MARL)-based scheme, which offers an effective strategy to tackle the challenges of high-dimensional signal processing. In the section of numerical results, we compare various algorithms with different network architectures. These comparisons reveal that the proposed MARL-based cooperative architecture can effectively strike a balance between system performance and communication overhead, thereby improving energy efficiency performance. It is important to note that increasing the number of user equipments participating in information sharing can effectively enhance SE performance, which also leads to an increase in power consumption, resulting in a non-trivial trade-off between the number of participants and EE performance.

Silicon photomultipliers (SiPMs) are being considered as potential replacements for conventional photomultiplier tubes (PMTs). However, a significant disadvantage of SiPMs is crosstalk (CT), wherein photons propagate through other pixels, resulting in secondary avalanches. CT can be categorized into internal crosstalk and external crosstalk based on whether the secondary avalanche occurs within the same SiPM or a different one. Numerous methods exist for quantitatively estimating the percentage of internal crosstalk (iCT). However, external crosstalk (eCT) has not been extensively studied. This article presents a novel measurement method for the probability of emitting an external crosstalk photon during a single pixel avalanche, using a setup involving two identical SiPMs facing each other, and without the need for complex optical designs. The entire apparatus is enclosed within a stainless steel chamber, functioning as a light-tight enclosure, and maintained at liquid nitrogen temperature. The experimental setup incorporates two Sensl J-60035 SiPM chips along with two 0.5-inch Hamamatsu Photonics (HPK) VUV4 S13370-6050CN SiPM arrays. The findings show a linear relationship between the probability of emitting an external crosstalk photon and the SiPM overvoltage for both SiPM samples. Surprisingly, this novel measurement method also rovides measurements of the SiPM photon detection efficiency (PDE) for eCT photons at low temperature.

Mastering a skill generally relies on both hands-on experience from doers and insightful, high-level guidance by mentors. Will this strategy also work well for solving complex non-convex optimization problems? Here, a common gradient-based optimizer acts like a disciplined doer, making locally optimal updates at each step. Large Language Models (LLMs) can also search for better solutions by inferring from natural language instructions, akin to a high-level mentor. In this paper, we show that these two participators are complementary to each other and can effectively collaborate as a combined optimization framework. The collaborative optimization is achieved by alternating between the gradient-based and LLM-based optimizers. We instruct LLMs to generate possibly improved solutions by taking parameter trajectories recorded during the previous stage of gradient-based optimization into account. Inferred results of LLMs are used as restarting points for the next stage of gradient optimization. We verify the effectiveness of this optimization framework on prompt tuning. By leveraging both the locally rigorous gradient-based optimizer and the high-level deductive LLM-based optimizer, the combined optimization method consistently yields improvements over competitive baselines on a variety of tasks. Our results demonstrate the synergistic effect of conventional gradient-based optimization and the inference ability of LLMs. The code is released at https://github.com/guozix/LLM-catalyst.

Directly generating material structures with optimal properties is a long-standing goal in material design. One of the fundamental challenges lies in how to overcome the limitation of traditional generative models to efficiently explore the global chemical space rather than a small localized space. Herein, we develop a framework named MAGECS to address this dilemma, by integrating the bird swarm algorithm and supervised graph neural network to effectively navigate the generative model in the immense chemical space towards materials with target properties. As a demonstration, MAGECS is applied to design compelling alloy electrocatalysts for CO$_2$ reduction reaction (CO$_2$RR) and works extremely well. Specifically, the chemical space of CO$_2$RR is effectively explored, where over 250,000 promising structures with high activity have been generated and notably, the proportion of desired structures is 2.5-fold increased. Moreover, five predicted alloys, i.e., CuAl, AlPd, Sn$_2$Pd$_5$, Sn$_9$Pd$_7$, and CuAlSe$_2$ are successfully synthesized and characterized experimentally, two of which exhibit about 90% Faraday efficiency of CO$_2$RR, and CuAl achieved 76% efficiency for C$_2$ products. This pioneering application of inverse design in CO$_2$RR catalysis showcases the potential of MAGECS to dramatically accelerate the development of functional materials, paving the way for fully automated, artificial intelligence-driven material design.

Large language models (LLMs) have catalyzed a paradigm shift in natural language processing, yet their limited controllability poses a significant challenge for downstream applications. We aim to address this by drawing inspiration from the neural mechanisms of the human brain, specifically Broca's and Wernicke's areas, which are crucial for language generation and comprehension, respectively. In particular, Broca's area receives cognitive decision signals from Wernicke's area, treating the language generation as an intricate decision-making process, which differs from the fully auto-regressive language generation of existing LLMs. In a similar vein, our proposed system, the BWArea model, conceptualizes language generation as a decision-making task. This model has three components: a language world model, an inverse dynamics model, and a cognitive policy. Like Wernicke's area, the inverse dynamics model is designed to deduce the underlying cognitive intentions, or latent actions, behind each token. The BWArea model is amenable to both pre-training and fine-tuning like existing LLMs. With 30B clean pre-training tokens, we have trained a BWArea model, which achieves competitive performance with LLMs of equal size (1B parameters). Unlike fully auto-regressive LLMs, its pre-training performance does not degenerate if dirty data unintentionally appears. This shows the advantage of a decomposed structure of BWArea model in reducing efforts in laborious data selection and labeling. Finally, we reveal that the BWArea model offers enhanced controllability via fine-tuning the cognitive policy with downstream reward metrics, thereby facilitating alignment with greater simplicity. On 9 out of 10 tasks from two suites, TextWorld and BigBench Hard, our method shows superior performance to auto-regressive LLMs.

Model-based methods in reinforcement learning offer a promising approach to enhance data efficiency by facilitating policy exploration within a dynamics model. However, accurately predicting sequential steps in the dynamics model remains a challenge due to the bootstrapping prediction, which attributes the next state to the prediction of the current state. This leads to accumulated errors during model roll-out. In this paper, we propose the Any-step Dynamics Model (ADM) to mitigate the compounding error by reducing bootstrapping prediction to direct prediction. ADM allows for the use of variable-length plans as inputs for predicting future states without frequent bootstrapping. We design two algorithms, ADMPO-ON and ADMPO-OFF, which apply ADM in online and offline model-based frameworks, respectively. In the online setting, ADMPO-ON demonstrates improved sample efficiency compared to previous state-of-the-art methods. In the offline setting, ADMPO-OFF not only demonstrates superior performance compared to recent state-of-the-art offline approaches but also offers better quantification of model uncertainty using only a single ADM.

The tool-use Large Language Models (LLMs) that integrate with external Python interpreters have significantly enhanced mathematical reasoning capabilities for open-source LLMs, while tool-free methods chose another track: augmenting math reasoning data. However, a great method to integrate the above two research paths and combine their advantages remains to be explored. In this work, we firstly include new math questions via multi-perspective data augmenting methods and then synthesize code-nested solutions to them. The open LLMs (i.e., Llama-2) are finetuned on the augmented dataset to get the resulting models, MuMath-Code ($\mu$-Math-Code). During the inference phase, our MuMath-Code generates code and interacts with the external python interpreter to get the execution results. Therefore, MuMath-Code leverages the advantages of both the external tool and data augmentation. To fully leverage the advantages of our augmented data, we propose a two-stage training strategy: In Stage-1, we finetune Llama-2 on pure CoT data to get an intermediate model, which then is trained on the code-nested data in Stage-2 to get the resulting MuMath-Code. Our MuMath-Code-7B achieves 83.8 on GSM8K and 52.4 on MATH, while MuMath-Code-70B model achieves new state-of-the-art performance among open methods -- achieving 90.7% on GSM8K and 55.1% on MATH. Extensive experiments validate the combination of tool use and data augmentation, as well as our two-stage training strategy. We release the proposed dataset along with the associated code for public use., Comment: The state-of-the-art open-source tool-use LLMs for mathematical reasoning

Text-to-image (T2I) diffusion models have shown significant success in personalized text-to-image generation, which aims to generate novel images with human identities indicated by the reference images. Despite promising identity fidelity has been achieved by several tuning-free methods, they usually suffer from overfitting issues. The learned identity tends to entangle with irrelevant information, resulting in unsatisfied text controllability, especially on faces. In this work, we present MasterWeaver, a test-time tuning-free method designed to generate personalized images with both faithful identity fidelity and flexible editability. Specifically, MasterWeaver adopts an encoder to extract identity features and steers the image generation through additional introduced cross attention. To improve editability while maintaining identity fidelity, we propose an editing direction loss for training, which aligns the editing directions of our MasterWeaver with those of the original T2I model. Additionally, a face-augmented dataset is constructed to facilitate disentangled identity learning, and further improve the editability. Extensive experiments demonstrate that our MasterWeaver can not only generate personalized images with faithful identity, but also exhibit superiority in text controllability. Our code can be found at https://github.com/csyxwei/MasterWeaver., Comment: ECCV 2024. Our code can be found at https://github.com/csyxwei/MasterWeaver