Search

Recent self-rewarding large language models (LLM) have successfully applied LLM-as-a-Judge to iteratively improve the alignment performance without the need of human annotations for preference data. These methods commonly utilize the same LLM to act as both the policy model (which generates responses) and the reward model (which scores and ranks those responses). The ranked responses are then used as preference pairs to train the LLM via direct alignment technologies (e.g. DPO). However, it is noteworthy that throughout this process, there is no guarantee of accuracy in the rewarding and ranking, which is critical for ensuring accurate rewards and high-quality preference data. Empirical results from relatively small LLMs (e.g., 7B parameters) also indicate that improvements from self-rewarding may diminish after several iterations in certain situations, which we hypothesize is due to accumulated bias in the reward system. This bias can lead to unreliable preference data for training the LLM. To address this issue, we first formulate and analyze the generalized iterative preference fine-tuning framework for self-rewarding language model. We then introduce the regularization to this generalized framework to mitigate the overconfident preference labeling in the self-rewarding process. Based on this theoretical insight, we propose a Consistency Regularized sElf-rewarding lAnguage Model (CREAM) that leverages the rewarding consistency across different iterations to regularize the self-rewarding training, helping the model to learn from more reliable preference data. With this explicit regularization, our empirical results demonstrate the superiority of CREAM in improving both reward consistency and alignment performance. The code is publicly available at https://github.com/Raibows/CREAM.

Interleaved multimodal comprehension and generation, enabling models to produce and interpret both images and text in arbitrary sequences, have become a pivotal area in multimodal learning. Despite significant advancements, the evaluation of this capability remains insufficient. Existing benchmarks suffer from limitations in data scale, scope, and evaluation depth, while current evaluation metrics are often costly or biased, lacking in reliability for practical applications. To address these challenges, we introduce MMIE, a large-scale knowledge-intensive benchmark for evaluating interleaved multimodal comprehension and generation in Large Vision-Language Models (LVLMs). MMIE comprises 20K meticulously curated multimodal queries, spanning 3 categories, 12 fields, and 102 subfields, including mathematics, coding, physics, literature, health, and arts. It supports both interleaved inputs and outputs, offering a mix of multiple-choice and open-ended question formats to evaluate diverse competencies. Moreover, we propose a reliable automated evaluation metric, leveraging a scoring model fine-tuned with human-annotated data and systematic evaluation criteria, aimed at reducing bias and improving evaluation accuracy. Extensive experiments demonstrate the effectiveness of our benchmark and metrics in providing a comprehensive evaluation of interleaved LVLMs. Specifically, we evaluate eight LVLMs, revealing that even the best models show significant room for improvement, with most achieving only moderate results. We believe MMIE will drive further advancements in the development of interleaved LVLMs. We publicly release our benchmark and code in https://mmie-bench.github.io/.

This work is concerned with fractional stochastic differential equations with different scales. We establish the existence and uniqueness of solutions for Caputo fractional stochastic differential systems under the non-Lipschitz condition. Based on the idea of temporal homogenization, we prove that the homogenization principle (averaging principle) holds in the sense of mean square ($L^2$ norm) convergence under a novel homogenization assumption. Furthermore, an Euler-Maruyama scheme for the non-autonomous system is constructed and its numerical error is analyzed. Finally, two numerical examples are presented to verify the theoretical results. Different from the existing literature, we demonstrate the computational advantages of the homogenized autonomous system from a numerical perspective.

Artificial intelligence has significantly impacted medical applications, particularly with the advent of Medical Large Vision Language Models (Med-LVLMs), sparking optimism for the future of automated and personalized healthcare. However, the trustworthiness of Med-LVLMs remains unverified, posing significant risks for future model deployment. In this paper, we introduce CARES and aim to comprehensively evaluate the Trustworthiness of Med-LVLMs across the medical domain. We assess the trustworthiness of Med-LVLMs across five dimensions, including trustfulness, fairness, safety, privacy, and robustness. CARES comprises about 41K question-answer pairs in both closed and open-ended formats, covering 16 medical image modalities and 27 anatomical regions. Our analysis reveals that the models consistently exhibit concerns regarding trustworthiness, often displaying factual inaccuracies and failing to maintain fairness across different demographic groups. Furthermore, they are vulnerable to attacks and demonstrate a lack of privacy awareness. We publicly release our benchmark and code in https://cares-ai.github.io/., Comment: NeurIPS 2024 Datasets and Benchmarks Track

Large vision-language models (LVLMs) have achieved impressive results in various visual question-answering and reasoning tasks through vision instruction tuning on specific datasets. However, there is still significant room for improvement in the alignment between visual and language modalities. Previous methods to enhance this alignment typically require external models or data, heavily depending on their capabilities and quality, which inevitably sets an upper bound on performance. In this paper, we propose SIMA, a framework that enhances visual and language modality alignment through self-improvement, eliminating the needs for external models or data. SIMA leverages prompts from existing vision instruction tuning datasets to self-generate responses and employs an in-context self-critic mechanism to select response pairs for preference tuning. The key innovation is the introduction of three vision metrics during the in-context self-critic process, which can guide the LVLM in selecting responses that enhance image comprehension. Through experiments across 14 hallucination and comprehensive benchmarks, we demonstrate that SIMA not only improves model performance across all benchmarks but also achieves superior modality alignment, outperforming previous approaches., Comment: 15 pages, 8 figures

Broadband spectrum detectors exhibit great promise in fields such as multispectral imaging and optical communications. Despite significant progress, challenges like materials instability, complex manufacturing process and high costs still hinder further application. Here we present a method that achieves broadband spectral detect by impurity-level in SrSnO3. We report over 200 mA/W photo-responsivity at 275 nm (ultraviolet C solar-bind) and 367 nm (ultraviolet A) and ~ 1 mA/W photo-responsivity at 532 nm and 700 nm (visible) with a voltage bias of 5V. Further transport and photoluminescence results indicate that the broadband response comes from the impurity levels and mutual interactions. Additionally, the photodetector demonstrates excellent robustness and stability under repeated tests and prolonged exposure in air. These findings show the potential of SSO photodetectors and propose a method to achieve broadband spectrum detection, creating new possibility for the development of single-phase, low-cost, simple structure and high-efficiency photodetectors., Comment: 5 Figures

Topological orders emerge in both microscopic quantum dynamics and macroscopic materials as a fundamental principle to characterize intricate properties in nature with vital significance, for instance, the Landau levels of electron systems in magnetic field. Whilst, recent advances of synthetic photonic systems enable generalized concepts of Landau levels across fermionic and bosonic systems, extending the modern physical frontier. However, the controls of Landau levels of photons were only confined in complex artificial metamaterials or multifolded cavities. Here, we exploit advanced structured light laser technology and propose the theory of high-dimensional frequency-degeneracy, which enables photonic Landau level control in a linear open laser cavity with simple displacement tuning of intracavity elements. This work not only create novel structured light with new topological effects but also provides broad prospects for Bose-analogue quantum Hall effects and topological physics.

Existing hyperspectral image (HSI) super-resolution (SR) methods struggle to effectively capture the complex spectral-spatial relationships and low-level details, while diffusion models represent a promising generative model known for their exceptional performance in modeling complex relations and learning high and low-level visual features. The direct application of diffusion models to HSI SR is hampered by challenges such as difficulties in model convergence and protracted inference time. In this work, we introduce a novel Group-Autoencoder (GAE) framework that synergistically combines with the diffusion model to construct a highly effective HSI SR model (DMGASR). Our proposed GAE framework encodes high-dimensional HSI data into low-dimensional latent space where the diffusion model works, thereby alleviating the difficulty of training the diffusion model while maintaining band correlation and considerably reducing inference time. Experimental results on both natural and remote sensing hyperspectral datasets demonstrate that the proposed method is superior to other state-of-the-art methods both visually and metrically., Comment: Accepted by AAAI2024

Existing free-energy guided No-Reference Image Quality Assessment (NR-IQA) methods still suffer from finding a balance between learning feature information at the pixel level of the image and capturing high-level feature information and the efficient utilization of the obtained high-level feature information remains a challenge. As a novel class of state-of-the-art (SOTA) generative model, the diffusion model exhibits the capability to model intricate relationships, enabling a comprehensive understanding of images and possessing a better learning of both high-level and low-level visual features. In view of these, we pioneer the exploration of the diffusion model into the domain of NR-IQA. Firstly, we devise a new diffusion restoration network that leverages the produced enhanced image and noise-containing images, incorporating nonlinear features obtained during the denoising process of the diffusion model, as high-level visual information. Secondly, two visual evaluation branches are designed to comprehensively analyze the obtained high-level feature information. These include the visual compensation guidance branch, grounded in the transformer architecture and noise embedding strategy, and the visual difference analysis branch, built on the ResNet architecture and the residual transposed attention block. Extensive experiments are conducted on seven public NR-IQA datasets, and the results demonstrate that the proposed model outperforms SOTA methods for NR-IQA.

In this paper, we propose a plasmon-induced transparency (PIT) metamaterial structure composed of Ag nanomaterials with polarization sensitivity. The metamaterial model consists of three bright modes with different resonant frequencies. The optical properties of the structure are further investigated using finite difference time domain (FDTD) method. The results show that the conversion between single-band PIT and dual-band PIT effects can be achieved by changing the polarization degree of the incident light, the number of transparent windows can be changed from one to two, and the process is accompanied by the conversion of bright and dark modes and the change of the resonance wavelength of the transmission peak. In addition, When the light is polarized in the Y-direction, the two transparency windows have different refractive index sensitivities, with FOM values of 5.94/RIU and 5.65/RIU, respectively.

Hypergraph Neural Networks (HGNNs) have been successfully applied in various hypergraph-related tasks due to their excellent higher-order representation capabilities. Recent works have shown that deep learning models are vulnerable to adversarial attacks. Most studies on graph adversarial attacks have focused on Graph Neural Networks (GNNs), and the study of adversarial attacks on HGNNs remains largely unexplored. In this paper, we try to reduce this gap. We design a new HGNNs attack model for the untargeted attack, namely MGHGA, which focuses on modifying node features. We consider the process of HGNNs training and use a surrogate model to implement the attack before hypergraph modeling. Specifically, MGHGA consists of two parts: feature selection and feature modification. We use a momentum gradient mechanism to choose the attack node features in the feature selection module. In the feature modification module, we use two feature generation approaches (direct modification and sign gradient) to enable MGHGA to be employed on discrete and continuous datasets. We conduct extensive experiments on five benchmark datasets to validate the attack performance of MGHGA in the node and the visual object classification tasks. The results show that MGHGA improves performance by an average of 2% compared to the than the baselines.

Large language models (LLMs) exhibit impressive emergent abilities in natural language processing, but their democratization is hindered due to huge computation requirements and closed-source nature. Recent research on advancing open-source smaller LMs by distilling knowledge from black-box LLMs has obtained promising results in the instruction-following ability. However, the reasoning ability which is more challenging to foster, is relatively rarely explored. In this paper, we propose a tailored learning approach to distill such reasoning ability to smaller LMs to facilitate the democratization of the exclusive reasoning ability. In contrast to merely employing LLM as a data annotator, we exploit the potential of LLM as a reasoning teacher by building an interactive multi-round learning paradigm. This paradigm enables the student to expose its deficiencies to the black-box teacher who then can provide customized training data in return. Further, to exploit the reasoning potential of the smaller LM, we propose self-reflection learning to motivate the student to learn from self-made mistakes. The learning from self-reflection and LLM are all tailored to the student's learning status, thanks to the seamless integration with the multi-round learning paradigm. Comprehensive experiments and analysis on mathematical and commonsense reasoning tasks demonstrate the effectiveness of our method. The code will be available at https://github.com/Raibows/Learn-to-Reason., Comment: To appear at EMNLP 2023

Rendering large adaptive mesh refinement (AMR) data in real-time in virtual reality (VR) environments is a complex challenge that demands sophisticated techniques and tools. The proposed solution harnesses the ExaBrick framework and integrates it as a plugin in COVISE, a robust visualization system equipped with the VR-centric OpenCOVER render module. This setup enables direct navigation and interaction within the rendered volume in a VR environment. The user interface incorporates rendering options and functions, ensuring a smooth and interactive experience. We show that high-quality volume rendering of AMR data in VR environments at interactive rates is possible using GPUs.

In this work, we propose a simple numerical scheme based on a fast front-tracking approach for solving a fluid-structure interaction (FSI) problem of plaque growth in blood vessels. A rigorous error analysis is carried out for the temporal semi-discrete scheme to show that it is first-order accurate for all macro time step $\Delta T$, micro time step $\Delta t$ and scale parameter $\epsilon$. A numerical example is presented to verify the theoretical results and demonstrate the excellent performance of the proposed multiscale algorithm.

Control of structured light is of great importance to explore fundamental physical effects and extend practical scientific applications, which has been advanced by accepting methods of quantum optics - many classical analogies of exotic quantum states were designed using structured modes. However, the prevailing quantum-like structured modes are limited by discrete states where the mode index is analog to the photon number state. Yet, beyond discrete states, there is a broad range of quantum states to be explored in the field of structured light -- continuous-variable (CV) states. As a typical example of CV states, squeezed state plays a prominent role in high-sensitivity interferometry and gravitational wave detection. In this work, we bring together two seemingly disparate branches of physics, namely, classical structured light and quantum squeezed state. We propose the structured light analogy of squeezed state (SLASS), which can break the spatial limit following the process of surpassing the standard quantum limit (SQL) with quantum squeezed states. This work paves the way for adopting methods from CV quantum states into structured light, opening new research directions of CV entanglement, teleportation, classical and quantum informatics of structured light in the future., Comment: 8 pages, 4 figures

This paper is concerned with a blood flow problem coupled with a slow plaque growth at the artery wall. In the model, the micro (fast) system is the Navier-Stokes equation with a periodically applied force and the macro (slow) system is a fractional reaction equation, which is used to describe the plaque growth with memory effect. We construct an auxiliary temporal periodic problem and an effective time-average equation to approximate the original problem and analyze the approximation error of the corresponding linearized PDE (Stokes) system, where the simple front-tracking technique is used to update the slow moving boundary. An effective multiscale method is then designed based on the approximate problem and the front tracking framework. We also present a temporal finite difference scheme with a spatial continuous finite element method and analyze its temporal discrete error. Furthermore, a fast iterative procedure is designed to find the initial value of the temporal periodic problem and its convergence is analyzed as well. Our designed front-tracking framework and the iterative procedure for solving the temporal periodic problem make it easy to implement the multiscale method on existing PDE solving software. The numerical method is implemented by a combination of the finite element platform COMSOL Multiphysics and the mainstream software MATLAB, which significantly reduce the programming effort and easily handle the fluid-structure interaction, especially moving boundaries with more complex geometries. We present some numerical examples of ODEs and 2-D Navier-Stokes system to demonstrate the effectiveness of the multiscale method. Finally, we have a numerical experiment on the plaque growth problem and discuss the physical implication of the fractional order parameter.