Search

Recently, numerous new benchmarks have been established to evaluate the performance of large language models (LLMs) via either computing a holistic score or employing another LLM as a judge. However, these approaches suffer from data leakage due to the open access of the benchmark and inflexible evaluation process. To address this issue, we introduce $\textbf{TreeEval}$, a benchmark-free evaluation method for LLMs that let a high-performance LLM host an irreproducible evaluation session and essentially avoids the data leakage. Moreover, this LLM performs as an examiner to raise up a series of questions under a topic with a tree planing strategy, which considers the current evaluation status to decide the next question generation and ensures the completeness and efficiency of the evaluation process. We evaluate $6$ models of different parameter sizes, including $7$B, $13$B, and $33$B, and ultimately achieved the highest correlation coefficient with AlpacaEval2.0 using only around $45$ questions. We also conduct more analysis to show the robustness and reliability of TreeEval. Our code can be accessed via the provided https://github.com/Ashura5/TreeEval.

Large language models (LLMs) undergo safety alignment to ensure safe conversations with humans. However, this paper introduces a training-free attack method capable of reversing safety alignment, converting the outcomes of stronger alignment into greater potential for harm by accessing only LLM output token distributions. Specifically, our method achieves this reversal by contrasting the output token distribution of a safety-aligned language model (e.g., Llama-2-chat) against its pre-trained version (e.g., Llama-2), so that the token predictions are shifted towards the opposite direction of safety alignment. We name this method emulated disalignment (ED) because sampling from this contrastive distribution provably emulates the result of fine-tuning to minimize a safety reward. Our experiments with ED across three evaluation datasets and four model families (Llama-1, Llama-2, Mistral, and Alpaca) show that ED doubles the harmfulness of pre-trained models and outperforms strong baselines, achieving the highest harmful rates in 43 out of 48 evaluation subsets by a large margin. Eventually, given ED's reliance on language model output token distributions, which particularly compromises open-source models, our findings highlight the need to reassess the open accessibility of language models, even if they have been safety-aligned. Code is available at https://github.com/ZHZisZZ/emulated-disalignment., Comment: ACL 2024

Surrogate modeling is of great practical significance for parametric differential equation systems. In contrast to classical numerical methods, using physics-informed deep learning methods to construct simulators for such systems is a promising direction due to its potential to handle high dimensionality, which requires minimizing a loss over a training set of random samples. However, the random samples introduce statistical errors, which may become the dominant errors for the approximation of low-regularity and high-dimensional problems. In this work, we present a deep adaptive sampling method for surrogate modeling ($\text{DAS}^2$), where we generalize the deep adaptive sampling (DAS) method [62] [Tang, Wan and Yang, 2023] to build surrogate models for low-regularity parametric differential equations. In the parametric setting, the residual loss function can be regarded as an unnormalized probability density function (PDF) of the spatial and parametric variables. This PDF is approximated by a deep generative model, from which new samples are generated and added to the training set. Since the new samples match the residual-induced distribution, the refined training set can further reduce the statistical error in the current approximate solution. We demonstrate the effectiveness of $\text{DAS}^2$ with a series of numerical experiments, including the parametric lid-driven 2D cavity flow problem with a continuous range of Reynolds numbers from 100 to 1000.

We present an efficient quantum circuit for block encoding pairing Hamiltonian often studied in nuclear physics. Our block encoding scheme does not require mapping the creation and annihilation operators to the Pauli operators and representing the Hamiltonian as a linear combination of unitaries. Instead, we show how to encode the Hamiltonian directly using controlled swap operations. We analyze the gate complexity of the block encoding circuit and show that it scales polynomially with respect to the number of qubits required to represent a quantum state associated with the pairing Hamiltonian. We also show how the block encoding circuit can be combined with the quantum singular value transformation to construct an efficient quantum circuit for approximating the density of states of a pairing Hamiltonian. The techniques presented can be extended to encode more general second-quantized Hamiltonians., Comment: 27 pages, 18 figures

Large Language Models (LLMs) are now commonplace in conversation applications. However, their risks of misuse for generating harmful responses have raised serious societal concerns and spurred recent research on LLM conversation safety. Therefore, in this survey, we provide a comprehensive overview of recent studies, covering three critical aspects of LLM conversation safety: attacks, defenses, and evaluations. Our goal is to provide a structured summary that enhances understanding of LLM conversation safety and encourages further investigation into this important subject. For easy reference, we have categorized all the studies mentioned in this survey according to our taxonomy, available at: https://github.com/niconi19/LLM-conversation-safety., Comment: Accepted to NAACL 2024

Recent advances in large language models (LLMs) have stepped forward the development of multilingual speech and machine translation by its reduced representation errors and incorporated external knowledge. However, both translation tasks typically utilize beam search decoding and top-1 hypothesis selection for inference. These techniques struggle to fully exploit the rich information in the diverse N-best hypotheses, making them less optimal for translation tasks that require a single, high-quality output sequence. In this paper, we propose a new generative paradigm for translation tasks, namely "GenTranslate", which builds upon LLMs to generate better results from the diverse translation versions in N-best list. Leveraging the rich linguistic knowledge and strong reasoning abilities of LLMs, our new paradigm can integrate the rich information in N-best candidates to generate a higher-quality translation result. Furthermore, to support LLM finetuning, we build and release a HypoTranslate dataset that contains over 592K hypotheses-translation pairs in 11 languages. Experiments on various speech and machine translation benchmarks (e.g., FLEURS, CoVoST-2, WMT) demonstrate that our GenTranslate significantly outperforms the state-of-the-art model., Comment: 18 pages, Accepted by ACL 2024. This work is open sourced at: https://github.com/YUCHEN005/GenTranslate

Recent studies have successfully shown that large language models (LLMs) can be successfully used for generative error correction (GER) on top of the automatic speech recognition (ASR) output. Specifically, an LLM is utilized to carry out a direct mapping from the N-best hypotheses list generated by an ASR system to the predicted output transcription. However, despite its effectiveness, GER introduces extra data uncertainty since the LLM is trained without taking into account acoustic information available in the speech signal. In this work, we aim to overcome such a limitation by infusing acoustic information before generating the predicted transcription through a novel late fusion solution termed Uncertainty-Aware Dynamic Fusion (UADF). UADF is a multimodal fusion approach implemented into an auto-regressive decoding process and works in two stages: (i) It first analyzes and calibrates the token-level LLM decision, and (ii) it then dynamically assimilates the information from the acoustic modality. Experimental evidence collected from various ASR tasks shows that UADF surpasses existing fusion mechanisms in several ways. It yields significant improvements in word error rate (WER) while mitigating data uncertainty issues in LLM and addressing the poor generalization relied with sole modality during fusion. We also demonstrate that UADF seamlessly adapts to audio-visual speech recognition., Comment: Accepted to ICLR 2024, 17 pages. This work will be open sourced under MIT license

A filter for inertial-based odometry is a recursive method used to estimate the pose from measurements of ego-motion and relative pose. Currently, there is no known filter that guarantees the computation of a globally optimal solution for the non-linear measurement model. In this paper, we demonstrate that an innovative filter, with the state being $SE_2(3)$ and the $\sqrt{n}$-\textit{consistent} pose as the initialization, efficiently achieves \textit{asymptotic optimality} in terms of minimum mean square error. This approach is tailored for real-time SLAM and inertial-based odometry applications. Our first contribution is that we propose an iterative filtering method based on the Gauss-Newton method on Lie groups which is numerically to solve the estimation of states from a priori and non-linear measurements. The filtering stands out due to its iterative mechanism and adaptive initialization. Second, when dealing with environmental measurements of the surroundings, we utilize a $\sqrt{n}$-consistent pose as the initial value for the update step in a single iteration. The solution is closed in form and has computational complexity $O(n)$. Third, we theoretically show that the approach can achieve asymptotic optimality in the sense of minimum mean square error from the a priori and virtual relative pose measurements (see Problem~\ref{prob:new update problem}). Finally, to validate our method, we carry out extensive numerical and experimental evaluations. Our results consistently demonstrate that our approach outperforms other state-of-the-art filter-based methods, including the iterated extended Kalman filter and the invariant extended Kalman filter, in terms of accuracy and running time.

In this paper, we show that the friends-and-strangers problem is PSPACE-complete by reduction from the Ncl (non-deterministic constraint logic) problem.

In high-stakes systems such as healthcare, it is critical to understand the causal reasons behind unusual events, such as sudden changes in patient's health. Unveiling the causal reasons helps with quick diagnoses and precise treatment planning. In this paper, we propose an automated method for uncovering "if-then" logic rules to explain observational events. We introduce temporal point processes to model the events of interest, and discover the set of latent rules to explain the occurrence of events. To achieve this, we employ an Expectation-Maximization (EM) algorithm. In the E-step, we calculate the likelihood of each event being explained by each discovered rule. In the M-step, we update both the rule set and model parameters to enhance the likelihood function's lower bound. Notably, we optimize the rule set in a differential manner. Our approach demonstrates accurate performance in both discovering rules and identifying root causes. We showcase its promising results using synthetic and real healthcare datasets., Comment: Accepted by AISTATS 2024

Attracted by the impressive power of Multimodal Large Language Models (MLLMs), the public is increasingly utilizing them to improve the efficiency of daily work. Nonetheless, the vulnerabilities of MLLMs to unsafe instructions bring huge safety risks when these models are deployed in real-world scenarios. In this paper, we systematically survey current efforts on the evaluation, attack, and defense of MLLMs' safety on images and text. We begin with introducing the overview of MLLMs on images and text and understanding of safety, which helps researchers know the detailed scope of our survey. Then, we review the evaluation datasets and metrics for measuring the safety of MLLMs. Next, we comprehensively present attack and defense techniques related to MLLMs' safety. Finally, we analyze several unsolved issues and discuss promising research directions. The latest papers are continually collected at https://github.com/isXinLiu/MLLM-Safety-Collection., Comment: Accepted at IJCAI2024

Fast and reliable transmission network reconfiguration is critical in improving power grid resilience to cyber-attacks. If the network reconfiguration following cyber-attacks is imperfect, secondary incidents may delay or interrupt post-attack restoration of the power grid. This paper proposes a framework of resilient transmission network reconfiguration, taking into account the impacts of cyber-attacks in the network reconfiguration process. First, the mechanism of cyber-attack propagation is analyzed based on the characteristics of network reconfiguration. Second, systematic resilience indices are specially extracted in which the impact of cyber-attacks on network reconfiguration is quantified. These indices are defined in terms of the restoration characteristics of the transmission power system. Third, representative cyber-attack incidents motivate an optimization-based model of resilient transmission network reconfiguration, and an optimal reconstruction scheme is obtained. Finally, simulation results based on the IEEE 39-bus system verify the feasibility and effectiveness of the proposed framework in enhancing power grid resilience to cyber-attacks.

Elucidating the reasoning process with structured explanations from question to answer is crucial, as it significantly enhances the interpretability, traceability, and trustworthiness of question-answering (QA) systems. However, structured explanations demand models to perform intricately structured reasoning, which poses great challenges. Most existing methods focus on single-step reasoning through supervised learning, ignoring logical dependencies between steps. Moreover, existing reinforcement learning (RL) based methods overlook the structured relationships, underutilizing the potential of RL in structured reasoning. In this paper, we propose SEER, a novel method that maximizes a structure-based return to facilitate structured reasoning and explanation. Our proposed structure-based return precisely describes the hierarchical and branching structure inherent in structured reasoning, effectively capturing the intricate relationships between different reasoning steps. In addition, we introduce a fine-grained reward function to meticulously delineate diverse reasoning steps. Extensive experiments show that SEER significantly outperforms state-of-the-art methods, achieving an absolute improvement of 6.9% over RL-based methods on EntailmentBank, a 4.4% average improvement on STREET benchmark, and exhibiting outstanding efficiency and cross-dataset generalization performance. Our code is available at https://github.com/Chen-GX/SEER., Comment: Camera ready version for ACL 2024 Main Conference

The detection of GW170817/GRB170817A implied the strong association between short gamma-ray bursts (SGRBs) and binary neutron star (BNS) mergers which produce gravitational waves (GWs). More evidence is needed to confirm the association and reveal the physical processes of BNS mergers. The upcoming High Altitude Detection of Astronomical Radiation (HADAR) experiment, excelling in a wide field of view (FOV) and a large effective area above tens of GeV, is a hope for the prompt detection of very-high-energy (VHE; > 10 GeV) SGRBs. The aim of this paper is to simulate and analyse GW/SGRB joint detections by future GW detector networks in synergy with HADAR, including the second generation LIGO, Virgo and KAGRA and the third generation ET and CE. We provide a brief introduction of the HADAR experiment for SGRB simulations and its expected SGRB detections. For GW simulations, we adopt a phenomenological model to describe GWs produced by BNS mergers and introduce the signal-noise ratios (SNRs) as detector responses. Following a theoretical analysis we compute the redshift-dependent efficiency functions of GW detector networks. We then construct the simulation of GW detection by Monte Carlo sampling. We compare the simulated results of LIGO-Virgo O2 and O3 runs with their actual detections as a check. The combination of GW and SGRB models is then discussed for joint detection, including parameter correlations, triggered SNRs and efficiency skymaps. The estimated joint detection rates are 0.09-2.52 per year for LHVK network with HADAR under different possible configurations, and approximately 0.27-7.89 per year for ET+CE network with HADAR.

Level statistics is a crucial tool in the exploration of localization physics. The level spacing distribution of the disordered localized phase follows Poisson statistics, and many studies naturally apply it to the quasiperiodic localized phase. Here we analytically obtain the level spacing distribution of the quasiperiodic localized phase, and find that it deviates from Poisson statistics. Moreover, based on this level statistics, we derive the ratio of adjacent gaps and find that for a single sample, it is a $\delta$ function, which is in excellent agreement with numerical studies. Additionally, unlike disordered systems, in quasiperiodic systems, there are variations in the level spacing distribution across different regions of the spectrum, and increasing the size and increasing the sample are non-equivalent. Our findings carry significant implications for the reevaluation of level statistics in quasiperiodic systems and a profound understanding of the distinct effects of quasiperiodic potentials and disorder induced localization.

Recent advances in large language models (LLMs) have promoted generative error correction (GER) for automatic speech recognition (ASR), which leverages the rich linguistic knowledge and powerful reasoning ability of LLMs to improve recognition results. The latest work proposes a GER benchmark with HyPoradise dataset to learn the mapping from ASR N-best hypotheses to ground-truth transcription by efficient LLM finetuning, which shows great effectiveness but lacks specificity on noise-robust ASR. In this work, we extend the benchmark to noisy conditions and investigate if we can teach LLMs to perform denoising for GER just like what robust ASR do}, where one solution is introducing noise information as a conditioner into LLM. However, directly incorporating noise embeddings from audio encoder could harm the LLM tuning due to cross-modality gap. To this end, we propose to extract a language-space noise embedding from the N-best list to represent the noise conditions of source speech, which can promote the denoising process in GER. Furthermore, in order to enhance its representation ability of audio noise, we design a knowledge distillation (KD) approach via mutual information estimation to distill the real noise information in audio embeddings to our language embedding. Experiments on various latest LLMs demonstrate our approach achieves a new breakthrough with up to 53.9% correction improvement in terms of word error rate while with limited training data. Analysis shows that our language-space noise embedding can well represent the noise conditions of source speech, under which off-the-shelf LLMs show strong ability of language-space denoising., Comment: Accepted to ICLR 2024, Spotlight top 5%, 24 pages. This work will be open sourced at: https://github.com/YUCHEN005/RobustGER under MIT license

The use of low-rank adaptation (LoRA) with frozen pretrained language models (PLMs) has become increasing popular as a mainstream, resource-efficient modeling approach for memory-constrained hardware. In this study, we first explore how to enhance model performance by introducing various LoRA training strategies, achieving relative word error rate reductions of 3.50\% on the public Librispeech dataset and of 3.67\% on an internal dataset in the messaging domain. To further characterize the stability of LoRA-based second-pass speech recognition models, we examine robustness against input perturbations. These perturbations are rooted in homophone replacements and a novel metric called N-best Perturbation-based Rescoring Robustness (NPRR), both designed to measure the relative degradation in the performance of rescoring models. Our experimental results indicate that while advanced variants of LoRA, such as dynamic rank-allocated LoRA, lead to performance degradation in $1$-best perturbation, they alleviate the degradation in $N$-best perturbation. This finding is in comparison to fully-tuned models and vanilla LoRA tuning baselines, suggesting that a comprehensive selection is needed when using LoRA-based adaptation for compute-cost savings and robust language modeling.

The streaming instability, a promising mechanism to drive planetesimal formation in dusty protoplanetary discs, relies on aerodynamic drag naturally induced by the background radial pressure gradient. This gradient should vary in disks, but its effect on the streaming instability has not been sufficiently explored. For this purpose, we use numerical simulations of an unstratified disc to study the non-linear saturation of the streaming instability with mono-disperse dust particles and survey a wide range of gradients for two distinct combinations of the particle stopping time and the dust-to-gas mass ratio. As the gradient increases, we find most kinematic and morphological properties increase but not always in linear proportion. The density distributions of tightly-coupled particles are insensitive to the gradient whereas marginally-coupled particles tend to concentrate by more than an order of magnitude as the gradient decreases. Moreover, dust-gas vortices for tightly-coupled particles shrink as the gradient decreases, and we note higher resolutions are required to trigger the instability in this case. In addition, we find various properties at saturation that depend on the gradient may be observable and may help reconstruct models of observed discs dominated by streaming turbulence. In general, increased dust diffusion from stronger gradients can lower the concentration of dust filaments and can explain the higher solid abundances needed to trigger strong particle clumping and the reduced planetesimal formation efficiency previously found in vertically-stratified simulations., Comment: Accepted by MNRAS. 22 pages, 15 figures, 5 tables

Kuiper belt objects show an unexpected trend, whereby large bodies have increasingly higher densities, up to five times greater than their smaller counterparts. Current explanations for this trend assume formation at constant composition, with the increasing density resulting from gravitational compaction. However, this scenario poses a timing problem to avoid early melting by decay of $^{26}$Al. We aim to explain the density trend in the context of streaming instability and pebble accretion. Small pebbles experience lofting into the atmosphere of the disk, being exposed to UV and partially losing their ice via desorption. Conversely, larger pebbles are shielded and remain more icy. We use a shearing box model including gas and solids, the latter split into ices and silicate pebbles. Self-gravity is included, allowing dense clumps to collapse into planetesimals. We find that the streaming instability leads to the formation of mostly icy planetesimals, albeit with an unexpected trend that the lighter ones are more silicate-rich than the heavier ones. We feed the resulting planetesimals into a pebble accretion integrator with a continuous size distribution, finding that they undergo drastic changes in composition as they preferentially accrete silicate pebbles. The density and masses of large KBOs are best reproduced if they form between 15 and 22\,AU. Our solution avoids the timing problem because the first planetesimals are primarily icy, and $^{26}$Al is mostly incorporated in the slow phase of silicate pebble accretion. Our results lend further credibility to the streaming instability and pebble accretion as formation and growth mechanisms., Comment: 24 pages, 13 figures, accepted to The Planetary Science Journal

Dust attenuation in star-forming galaxies (SFGs), as parameterized by the infrared excess (IRX $\equiv L_{\rm IR}/L_{\rm UV}$), is found to be tightly correlated with star formation rate (SFR), metallicity and galaxy size, following a universal IRX relation up to $z=3$. This scaling relation can provide a fundamental constraint for theoretical models to reconcile galaxy star formation, chemical enrichment, and structural evolution across cosmic time. We attempt to reproduce the universal IRX relation over $0.1\leq z\leq 2.5$ using the EAGLE hydrodynamical simulations and examine sensitive parameters in determining galaxy dust attenuation. Our findings show that while the predicted universal IRX relation from EAGLE approximately aligns with observations at $z\leq 0.5$, noticeable disparities arise at different stellar masses and higher redshifts. Specifically, we investigate how modifying various galaxy parameters can affect the predicted universal IRX relation in comparison to the observed data. We demonstrate that the simulated gas-phase metallicity is the critical quantity for the shape of the predicted universal IRX relation. We find that the influence of the infrared luminosity and infrared excess is less important while galaxy size has virtually no significant effect. Overall, the EAGLE simulations are not able to replicate some of the observed characteristics between IRX and galaxy parameters of SFGs, emphasizing the need for further investigation and testing for our current state-of-the-art theoretical models., Comment: 19 pages, 15 figures, accepted for publication in MNRAS

Large Language Models (LLMs) have demonstrated superior abilities in tasks such as chatting, reasoning, and question-answering. However, standard LLMs may ignore crucial paralinguistic information, such as sentiment, emotion, and speaking style, which are essential for achieving natural, human-like spoken conversation, especially when such information is conveyed by acoustic cues. We therefore propose Paralinguistics-enhanced Generative Pretrained Transformer (ParalinGPT), an LLM that utilizes text and speech modalities to better model the linguistic content and paralinguistic attributes of spoken dialogue. The model takes the conversational context of text, speech embeddings, and paralinguistic attributes as input prompts within a serialized multitasking multimodal framework. Specifically, our framework serializes tasks in the order of current paralinguistic attribute prediction, response paralinguistic attribute prediction, and response text generation with autoregressive conditioning. We utilize the Switchboard-1 corpus, including its sentiment labels as the paralinguistic attribute, as our spoken dialogue dataset. Experimental results indicate the proposed serialized multitasking method outperforms typical sequence classification techniques on current and response sentiment classification. Furthermore, leveraging conversational context and speech embeddings significantly improves both response text generation and sentiment prediction. Our proposed framework achieves relative improvements of 6.7%, 12.0%, and 3.5% in current sentiment accuracy, response sentiment accuracy, and response text BLEU score, respectively., Comment: Accepted by ICASSP 2024. Camera-ready version

Training large foundation models using self-supervised objectives on unlabeled data, followed by fine-tuning on downstream tasks, has emerged as a standard procedure. Unfortunately, the efficacy of this approach is often constrained by both limited fine-tuning compute and scarcity in labeled downstream data. We introduce Multimodal Attention Merging (MAM), an attempt that facilitates direct knowledge transfer from attention matrices of models rooted in high resource modalities, text and images, to those in resource-constrained domains, speech and audio, employing a zero-shot paradigm. MAM reduces the relative Word Error Rate (WER) of an Automatic Speech Recognition (ASR) model by up to 6.70%, and relative classification error of an Audio Event Classification (AEC) model by 10.63%. In cases where some data/compute is available, we present Learnable-MAM, a data-driven approach to merging attention matrices, resulting in a further 2.90% relative reduction in WER for ASR and 18.42% relative reduction in AEC compared to fine-tuning., Comment: 5 pages, 1 figure, ICASSP 2024 Workshop on Self-supervision in Audio, Speech and Beyond

Recent advancements in offline reinforcement learning (RL) have underscored the capabilities of Return-Conditioned Supervised Learning (RCSL), a paradigm that learns the action distribution based on target returns for each state in a supervised manner. However, prevailing RCSL methods largely focus on deterministic trajectory modeling, disregarding stochastic state transitions and the diversity of future trajectory distributions. A fundamental challenge arises from the inconsistency between the sampled returns within individual trajectories and the expected returns across multiple trajectories. Fortunately, value-based methods offer a solution by leveraging a value function to approximate the expected returns, thereby addressing the inconsistency effectively. Building upon these insights, we propose a novel approach, termed the Critic-Guided Decision Transformer (CGDT), which combines the predictability of long-term returns from value-based methods with the trajectory modeling capability of the Decision Transformer. By incorporating a learned value function, known as the critic, CGDT ensures a direct alignment between the specified target returns and the expected returns of actions. This integration bridges the gap between the deterministic nature of RCSL and the probabilistic characteristics of value-based methods. Empirical evaluations on stochastic environments and D4RL benchmark datasets demonstrate the superiority of CGDT over traditional RCSL methods. These results highlight the potential of CGDT to advance the state of the art in offline RL and extend the applicability of RCSL to a wide range of RL tasks., Comment: Accepted at AAAI 2024

The streaming instability (SI) is a leading candidate for planetesimal formation, which can concentrate solids through two-way aerodynamic interactions with the gas. The resulting concentrations can become sufficiently dense to collapse under particle self-gravity, forming planetesimals. Previous studies have carried out large parameter surveys to establish the critical particle to gas surface density ratio ($Z$), above which SI-induced concentration triggers planetesimal formation. The threshold $Z$ depends on the dimensionless stopping time ($\tau_s$, a proxy for dust size). However, these studies neglected both particle self-gravity and external turbulence. Here, we perform 3D stratified shearing box simulations with both particle self-gravity and turbulent forcing, which we characterize via $\alpha_D$ that measures turbulent diffusion. We find that forced turbulence, at amplitudes plausibly present in some protoplanetary disks, can increase the threshold $Z$ by up to an order of magnitude. For example, for $\tau_s = 0.01$, planetesimal formation occurs when $Z \gtrsim 0.06$, $\gtrsim 0.1$, and $\gtrsim 0.2$ at $\alpha_D = 10^{-4}$, $10^{-3.5}$, and $10^{-3}$, respectively. We provide a single fit to the critical $Z$ as a function of $\alpha_D$ and $\tau_s$ required for the SI to work (though limited to the range $\tau_s = 0.01$--0.1). Our simulations also show that planetesimal formation requires a mid-plane particle-to-gas density ratio that exceeds unity, with the critical value being independent of $\alpha_D$. Finally, we provide the estimation of particle scale height that accounts for both particle feedback and external turbulence., Comment: 27 pages, 13 figures, submitted to The Astrophysical Journal

When a Heisenberg honeycomb ferromagnet is irradiated by high frequency circularly polarized light, the underlying uncharged magnons acquire a time dependent Aharonov Casher phase, which makes it a Floquet topological magnon insulator. In this context, we investigate the many body interaction effects of Floquet magnons in laser irradiated Heisenberg honeycomb ferromagnets with ocontaining Dzyaloshinskii Moriya interaction under the application of circularly polarized off resonant light. We demonstrate that the quantum ferromagnet systems periodically laser driven exhibits temperature driven topological phase transitions due to Floquet magnon magnon interactions. The thermal Hall effect of Floquet magnons serves as a prominent signature for detecting these many body effects near the critical point, enabling experimental investigation into this phenomenon. Our study complements the lack of previous theoretical works that the topological phase transition of the Floquet magnon under the linear spin wave approximation is only tunable by the light field. Our study presents a novel approach for constructing Floquet topological phases in periodically driven quantum magnet systems that goes beyond the limitations of the linear spin wave theory. We provide numerical results based on the well known van der Waals quantum magnet CrX3 (X=F, Cl, Br, and I), calling for experimental implementation.

Aiming at improving the survey efficiency of the Wide Field Survey Telescope, we have developed a basic scheduling strategy that takes into account the telescope characteristics, observing conditions, and weather conditions at the Lenghu site. The sky area is divided into rectangular regions, referred to as `tiles', with a size of 2.577 deg * 2.634 deg slightly smaller than the focal area of the mosaic CCDs. These tiles are continuously filled in annulars parallel to the equator. The brightness of the sky background, which varies with the moon phase and distance from the moon, plays a significant role in determining the accessible survey fields. Approximately 50 connected tiles are grouped into one block for observation. To optimize the survey schedule, we perform simulations by taking into account the length of exposures, data readout, telescope slewing, and all relevant observing conditions. We utilize the Greedy Algorithm for scheduling optimization. Additionally, we propose a dedicated dithering pattern to cover the gaps between CCDs and the four corners of the mosaic CCD array, which are located outside of the 3 deg field of view. This dithering pattern helps to achieve relatively uniform exposure maps for the final survey outputs., Comment: 14 pages, 7 figures, 1 table. Accepted for pubulication in Research in Astronomy and Astrophysics

The security concerns surrounding Large Language Models (LLMs) have been extensively explored, yet the safety of Multimodal Large Language Models (MLLMs) remains understudied. In this paper, we observe that Multimodal Large Language Models (MLLMs) can be easily compromised by query-relevant images, as if the text query itself were malicious. To address this, we introduce MM-SafetyBench, a comprehensive framework designed for conducting safety-critical evaluations of MLLMs against such image-based manipulations. We have compiled a dataset comprising 13 scenarios, resulting in a total of 5,040 text-image pairs. Our analysis across 12 state-of-the-art models reveals that MLLMs are susceptible to breaches instigated by our approach, even when the equipped LLMs have been safety-aligned. In response, we propose a straightforward yet effective prompting strategy to enhance the resilience of MLLMs against these types of attacks. Our work underscores the need for a concerted effort to strengthen and enhance the safety measures of open-source MLLMs against potential malicious exploits. The resource is available at https://github.com/isXinLiu/MM-SafetyBench

The aim of video summarization is to shorten videos automatically while retaining the key information necessary to convey the overall story. Video summarization methods mainly rely on visual factors, such as visual consecutiveness and diversity, which may not be sufficient to fully understand the content of the video. There are other non-visual factors, such as interestingness, representativeness, and storyline consistency that should also be considered for generating high-quality video summaries. Current methods do not adequately take into account these non-visual factors, resulting in suboptimal performance. In this work, a new approach to video summarization is proposed based on insights gained from how humans create ground truth video summaries. The method utilizes a conditional modeling perspective and introduces multiple meaningful random variables and joint distributions to characterize the key components of video summarization. Helper distributions are employed to improve the training of the model. A conditional attention module is designed to mitigate potential performance degradation in the presence of multi-modal input. The proposed video summarization method incorporates the above innovative design choices that aim to narrow the gap between human-generated and machine-generated video summaries. Extensive experiments show that the proposed approach outperforms existing methods and achieves state-of-the-art performance on commonly used video summarization datasets., Comment: This work has been submitted to the IEEE for possible publication. arXiv admin note: substantial text overlap with arXiv:2305.00455

Nonequilibrium dynamics governed by electron-phonon (e-ph) interactions plays a key role in electronic devices and spectroscopies and is central to understanding electronic excitations in materials. The real-time Boltzmann transport equation (rt-BTE) with collision processes computed from first principles can describe the coupled dynamics of electrons and atomic vibrations (phonons). Yet, a bottleneck of these simulations is the calculation of e-ph scattering integrals on dense momentum grids at each time step. Here we show a data-driven approach based on dynamic mode decomposition (DMD) that can accelerate the time propagation of the rt-BTE and identify dominant electronic processes. We apply this approach to two case studies, high-field charge transport and ultrafast excited electron relaxation. In both cases, simulating only a short time window of ~10% of the dynamics suffices to predict the dynamics from initial excitation to steady state using DMD extrapolation. Analysis of the momentum-space modes extracted from DMD sheds light on the microscopic mechanisms governing electron relaxation to steady state or equilibrium. The combination of accuracy and efficiency makes our DMD-based method a valuable tool for investigating ultrafast dynamics in a wide range of materials.

The restoration control of the modern alternating current-direct current (AC-DC) hybrid power grid after a major blackout is difficult and complex. Taking into account the interaction between the line-commutated converter high-voltage direct current (LCC-HVDC) and the AC power grid, this paper proposes a novel optimization method of restoration path to reconfigure the skeleton network for the blackout power grid. Based on the system strength, the supporting capability of the AC power grid for the LCC-HVDC is first analysed from the aspects of start-up and operation of LCC-HVDCs. Subsequently, the quantitative relationship between the restoration path and the restoration characteristic of LCC-HVDC is derived in detail based on the system strength indices of the short-circuit capacity and the frequency regulation capability. Then, an optimization model of restoration path considering non-tree paths is formulated and a feasible optimization algorithm is proposed to achieve the optimal path restoration scheme. A modified IEEE 39-bus system and a partial power grid of Southwest China are simulated to show that the proposed method is suitable for the restoration of AC-DC power grids and can improve restoration efficiency. This research can be an important guidance for operators to rapidly restore the AC-DC power grid., Comment: Accepted by IET Generation, Transmission & Distribution