Search

We consider a $d$-dimensional stochastic differential equation (SDE) of the form $dU_t = b(U_t)d t + \sigma d Z_t$ with initial condition $U_0 = u\in \mathbb{R}^d$ and additive noise. Denote by $X_t^x$ the (unique) solution with initial value $X_0^x = x$, if the driving noise $Z_t$ is a $d$-dimensional rotationally symmetric $\alpha$-stable ($1<\alpha<2$) L\'evy process, and by $Y_t^y$ the (unique) solution with initial value $Y_0^y=y$, if the driving noise is a $d$-dimensional Brownian motion. We give precise estimates of the total variation distance $\|{\rm law} (X_{t}^x)-{\rm law} (Y_{t}^y)\|_{TV}$ for all $t>0$, and we show that the ergodic measures $\mu_\alpha$ and $\mu_2$ of $X_t^x$ and $Y_t^y$, respectively, satisfy $\|\mu_\alpha-\mu_2\|_{TV} \leq 2\wedge\left[\frac{Cd^{3/2}}{\alpha-1}(2-\alpha)\right]$. As an application, we consider the following $d$-dimensional Poisson PDE \begin{gather*} -(-\Delta)^{\alpha/2} f(x) + \langle b(x), \nabla f(x)\rangle = h(x)-\mu_\alpha(h), \end{gather*} where $h \in C_b(\mathbb{R}^d,\mathbb{R})$ and $1<\alpha\leq 2$. We show that this PDE admits a solution $f_\alpha$, which is continuous, but may have linear growth. For fixed $\alpha_0\in(1,2)$, we show that for any $\alpha_0\leq \alpha <2$ there is a constant $C(\alpha_0)$ such that \begin{gather*} \left|f_\alpha(x)-f_2(x)\right| \leq C(\alpha_0)(1+|x|)\|h\|_\infty d^{3/2}(2-\alpha)\log\frac{1}{2-\alpha} \quad\text{for all $x$}. \end{gather*}

Diffusion models have exhibited remarkable prowess in visual generalization. Building on this success, we introduce an instruction-based object addition pipeline, named Add-SD, which automatically inserts objects into realistic scenes with rational sizes and positions. Different from layout-conditioned methods, Add-SD is solely conditioned on simple text prompts rather than any other human-costly references like bounding boxes. Our work contributes in three aspects: proposing a dataset containing numerous instructed image pairs; fine-tuning a diffusion model for rational generation; and generating synthetic data to boost downstream tasks. The first aspect involves creating a RemovalDataset consisting of original-edited image pairs with textual instructions, where an object has been removed from the original image while maintaining strong pixel consistency in the background. These data pairs are then used for fine-tuning the Stable Diffusion (SD) model. Subsequently, the pretrained Add-SD model allows for the insertion of expected objects into an image with good rationale. Additionally, we generate synthetic instances for downstream task datasets at scale, particularly for tail classes, to alleviate the long-tailed problem. Downstream tasks benefit from the enriched dataset with enhanced diversity and rationale. Experiments on LVIS val demonstrate that Add-SD yields an improvement of 4.3 mAP on rare classes over the baseline. Code and models are available at https://github.com/ylingfeng/Add-SD.

Graph foundation models have recently attracted significant attention due to its strong generalizability. Although existing methods resort to language models to learn unified semantic representations across domains, they disregard the unique structural characteristics of graphs from different domains. To address the problem, in this paper, we boost graph foundation model from structural perspective and propose BooG. The model constructs virtual super nodes to unify structural characteristics of graph data from different domains. Specifically, the super nodes fuse the information of anchor nodes and class labels, where each anchor node captures the information of a node or a graph instance to be classified. Instead of using the raw graph structure, we connect super nodes to all nodes within their neighborhood by virtual edges. This new structure allows for effective information aggregation while unifying cross-domain structural characteristics. Additionally, we propose a novel pre-training objective based on contrastive learning, which learns more expressive representations for graph data and generalizes effectively to different domains and downstream tasks. Experimental results on various datasets and tasks demonstrate the superior performance of BooG. We provide our code and data here: https://anonymous.4open.science/r/BooG-EE42/.

Recommender system of the e-commerce platform usually serves multiple business scenarios. Multi-scenario Recommendation (MSR) is an important topic that improves ranking performance by leveraging information from different scenarios. Recent methods for MSR mostly construct scenario shared or specific modules to model commonalities and differences among scenarios. However, when the amount of data among scenarios is skewed or data in some scenarios is extremely sparse, it is difficult to learn scenario-specific parameters well. Besides, simple sharing of information from other scenarios may result in a negative transfer. In this paper, we propose a unified model named Cross-Scenario Information Interaction (CSII) to serve all scenarios by a mixture of scenario-dominated experts. Specifically, we propose a novel method to select highly transferable features in data instances. Then, we propose an attention-based aggregator module, which can adaptively extract relative knowledge from cross-scenario. Experiments on the production dataset verify the superiority of our method. Online A/B test in Meituan Waimai APP also shows a significant performance gain, leading to an average improvement in GMV (Gross Merchandise Value) of 1.0% for overall scenarios.

Understanding user interests is crucial for Click-Through Rate (CTR) prediction tasks. In sequential recommendation, pre-training from user historical behaviors through self-supervised learning can better comprehend user dynamic preferences, presenting the potential for direct integration with CTR tasks. Previous methods have integrated pre-trained models into downstream tasks with the sole purpose of extracting semantic information or well-represented user features, which are then incorporated as new features. However, these approaches tend to ignore the additional inference costs to the downstream tasks, and they do not consider how to transfer the effective information from the pre-trained models for specific estimated items in CTR prediction. In this paper, we propose a Sequential Recommendation Pre-training framework for CTR prediction (SRP4CTR) to tackle the above problems. Initially, we discuss the impact of introducing pre-trained models on inference costs. Subsequently, we introduced a pre-trained method to encode sequence side information concurrently.During the fine-tuning process, we incorporate a cross-attention block to establish a bridge between estimated items and the pre-trained model at a low cost. Moreover, we develop a querying transformer technique to facilitate the knowledge transfer from the pre-trained model to industrial CTR models. Offline and online experiments show that our method outperforms previous baseline models.

As data plays an increasingly pivotal role in decision-making, the emergence of data markets underscores the growing importance of data valuation. Within the machine learning landscape, Data Shapley stands out as a widely embraced method for data valuation. However, a limitation of Data Shapley is its assumption of a fixed dataset, contrasting with the dynamic nature of real-world applications where data constantly evolves and expands. This paper establishes the relationship between Data Shapley and infinite-order U-statistics and addresses this limitation by quantifying the uncertainty of Data Shapley with changes in data distribution from the perspective of U-statistics. We make statistical inferences on data valuation to obtain confidence intervals for the estimations. We construct two different algorithms to estimate this uncertainty and provide recommendations for their applicable situations. We also conduct a series of experiments on various datasets to verify asymptotic normality and propose a practical trading scenario enabled by this method.

4U 1954+31 is a high-mass X-ray binary (HMXB) that contains a neutron star and an M supergiant companion. The neutron star has a spin period of ~5.4 hr. The traditional wind-accreting model requires an ultra-strong magnetic field for the neutron star to explain its extremely long spin period, which seems problematic for the neutron star with an age of a few tens of million years. In this work, we take into account the unsteady feature of wind accretion, which results in alternation of the direction of the wind matter's angular momentum. Accordingly, the torque exerted by the accreted wind matter varies between positive and negative from time to time, and largely cancels out over long time. In such a scenario, neutron stars can naturally attain long spin periods without the requirement of a very strong magnetic field. This may also provide a reasonable explanation for the spin period distribution of long-period neutron stars in HMXBs., Comment: 9 pages, 6 figures, accepted for publication in MNRAS

GPT-4V has attracted considerable attention due to its extraordinary capacity for integrating and processing multimodal information. At the same time, its ability of face recognition raises new safety concerns of privacy leakage. Despite researchers' efforts in safety alignment through RLHF or preprocessing filters, vulnerabilities might still be exploited. In our study, we introduce AutoJailbreak, an innovative automatic jailbreak technique inspired by prompt optimization. We leverage Large Language Models (LLMs) for red-teaming to refine the jailbreak prompt and employ weak-to-strong in-context learning prompts to boost efficiency. Furthermore, we present an effective search method that incorporates early stopping to minimize optimization time and token expenditure. Our experiments demonstrate that AutoJailbreak significantly surpasses conventional methods, achieving an Attack Success Rate (ASR) exceeding 95.3\%. This research sheds light on strengthening GPT-4V security, underscoring the potential for LLMs to be exploited in compromising GPT-4V integrity., Comment: TrustNLP@NAACL2024 (Fourth Workshop on Trustworthy Natural Language Processing)

In this paper, we use the information-theoretic approach to study curvature-dimension condition, rigidity theorems and entropy differential inequalities on Riemannian manifolds. We prove the equivalence of the ${\rm CD}(K, m)$-condition for $K\in \mathbb{R}$ and $m\in [n, \infty]$ and some entropy differential inequalities along the geodesics on the Wasserstein space over a Riemannian manifold. The rigidity models of the entropy differential inequalities are the $K$-Einstein manifolds and the $(K, m)$-Einstein manifolds with Hessian solitons. Moreover, we prove the monotonicity and rigidity theorem of the $W$-entropy along the geodesics on the Wasserstein space over Riemannian manifolds with CD$(0, m)$-condition. Finally, we compare our work with the synthetic geometry developed by Lott-Villani and Sturm.

We study the radiation in kink collision via a model that varies between $\phi^6$ theory and $\phi^2$ theory with some discontinuities. Both numerical and analytical methods were used to investigate The kink-antikink(KAK) and antikink-kink(AKK) collision. In the numerical analysis, we found the critical velocities in both collisions increased with $n$. We also found a finite lifetime oscillon window in KAK collision for $n=2$. In the analytical part, we found a family of shock wave solutions that describes radiation in the kink collision perfectly. Moreover, an analytical AKK solution at $n\rightarrow\infty$ and $v=1$ was found by considering a certain limit of these solutions., Comment: 27 pages, 68 figures

Short fixed-length inputs are the main bottleneck of deep learning methods in long time-series forecasting tasks. Prolonging input length causes overfitting, rapidly deteriorating accuracy. Our research indicates that the overfitting is a combination reaction of the multi-scale pattern coupling in time series and the fixed focusing scale of current models. First, we find that the patterns exhibited by a time series across various scales are reflective of its multi-periodic nature, where each scale corresponds to specific period length. Second, We find that the token size predominantly dictates model behavior, as it determines the scale at which the model focuses and the context size it can accommodate. Our idea is to decouple the multi-scale temporal patterns of time series and to model each pattern with its corresponding period length as token size. We introduced a novel series-decomposition module(MPSD), and a Multi-Token Pattern Recognition neural network(MTPR), enabling the model to handle \textit{inputs up to $10\times$ longer}. Sufficient context enhances performance(\textit{38% maximum precision improvement}), and the decoupling approach offers \textit{Low complexity($0.22\times$ cost)} and \textit{high interpretability}., Comment: 9 pages

Providing natural language-based explanations to justify recommendations helps to improve users' satisfaction and gain users' trust. However, as current explanation generation methods are commonly trained with an objective to mimic existing user reviews, the generated explanations are often not aligned with the predicted ratings or some important features of the recommended items, and thus, are suboptimal in helping users make informed decision on the recommendation platform. To tackle this problem, we propose a flexible model-agnostic method named MMI (Maximizing Mutual Information) framework to enhance the alignment between the generated natural language explanations and the predicted rating/important item features. Specifically, we propose to use mutual information (MI) as a measure for the alignment and train a neural MI estimator. Then, we treat a well-trained explanation generation model as the backbone model and further fine-tune it through reinforcement learning with guidance from the MI estimator, which rewards a generated explanation that is more aligned with the predicted rating or a pre-defined feature of the recommended item. Experiments on three datasets demonstrate that our MMI framework can boost different backbone models, enabling them to outperform existing baselines in terms of alignment with predicted ratings and item features. Additionally, user studies verify that MI-enhanced explanations indeed facilitate users' decisions and are favorable compared with other baselines due to their better alignment properties., Comment: This paper has been accepted by cikm2024, and the code repository will be updated soon

Generative image reconstruction algorithms such as measurement conditioned diffusion models are increasingly popular in the field of medical imaging. These powerful models can transform low signal-to-noise ratio (SNR) inputs into outputs with the appearance of high SNR. However, the outputs can have a new type of error called hallucinations. In medical imaging, these hallucinations may not be obvious to a Radiologist but could cause diagnostic errors. Generally, hallucination refers to error in estimation of object structure caused by a machine learning model, but there is no widely accepted method to evaluate hallucination magnitude. In this work, we propose a new image quality metric called the hallucination index. Our approach is to compute the Hellinger distance from the distribution of reconstructed images to a zero hallucination reference distribution. To evaluate our approach, we conducted a numerical experiment with electron microscopy images, simulated noisy measurements, and applied diffusion based reconstructions. We sampled the measurements and the generative reconstructions repeatedly to compute the sample mean and covariance. For the zero hallucination reference, we used the forward diffusion process applied to ground truth. Our results show that higher measurement SNR leads to lower hallucination index for the same apparent image quality. We also evaluated the impact of early stopping in the reverse diffusion process and found that more modest denoising strengths can reduce hallucination. We believe this metric could be useful for evaluation of generative image reconstructions or as a warning label to inform radiologists about the degree of hallucinations in medical images.

Drug repositioning is an economically efficient strategy used to discover new indications for existing drugs beyond their original approvals, expanding their applicability and usage to address challenges in disease treatment. In recent years, deep-learning techniques for drug repositioning have gained much attention. While most deep learning-based research methods focus on encoding drugs and diseases by extracting feature information from neighbors in the network, they often pay little attention to the potential relationships between the features of drugs and diseases, leading to imprecise encoding of drugs and diseases. To address this, we design a dual-feature drug repositioning neural network (DFDRNN) model to achieve precise encoding of drugs and diseases. DFDRNN uses two features to represent drugs and diseases: the similarity feature and the association feature. The model incorporates a self-attention mechanism to design two dual-feature extraction modules for achieving precisely encoding of drugs and diseases: the intra-domain dual-feature extraction (IntraDDFE) module and the inter-domain dual-feature extraction (InterDDFE) module. The IntraDDFE module extracts features from a single domain (drug or disease domain), while the InterDDFE module extracts features from the mixed domain (drug and disease domain). In particular, the feature is changed by InterDDFE, ensuring a precise encoding of drugs and diseases. Finally, a cross-dual-domain decoder is designed to predict drug-disease associations in both the drug and disease domains. Compared to six state-of-the-art methods, DFDRNN outperforms others on four benchmark datasets, with an average AUROC of 0.946 and an average AUPR of 0.597.

As the largest gravitationally bound objects in the Universe, galaxy clusters have provided the first piece of evidence for the presence of dark matter and may be suitable targets for indirect dark matter searches. Among various signals, the GeV-TeV $\gamma$-ray line has been taken as the smoking-gun signal of the dark matter annihilation/decay since no known astrophysical/physical process(es) could generate such a peculiar spectrum. With 15.5 years of Fermi-LAT P8R3 publicly available data, we search for the $\gamma$-ray line emission in the directions of a group of 13 nearby massive galaxy clusters with an unbinned likelihood analysis. A $\gamma$-ray line signal at $\sim 43.2$ GeV has a net TS value of $\approx 30$ if we only take into account the data in the directions of Virgo, Fornax and Ophiuchus clusters, three massive clusters with the highest J-factors expected to generate the dark matter annihilation signal. The signal still presents when the data of other 10 nearby massive clusters have also been included, though the TS value decreases to $\approx 21$ likely because of their lower signal-to-noise ratios. The absence of this signal in the inner Galaxy disfavors both the instrumental effect and the canonical dark matter annihilation interpretation, and a more sophisticated dark matter model or very peculiar astrophysical scenario might be needed. This $\gamma$-ray line signal, if intrinsic, could be unambiguously verified by the Very Large Area $\gamma$-ray Space Telescope in its first two years of performance., Comment: 9 pages, 7 figures, 1 table. Comments are welcome!

The matrix product state (MPS) ansatz offers a promising approach for finding the ground state of molecular Hamiltonians and solving quantum chemistry problems. Building on this concept, the proposed technique of quantum circuit MPS (QCMPS) enables the simulation of chemical systems using a relatively small number of qubits. In this study, we enhance the optimization performance of the QCMPS ansatz by employing the variational quantum imaginary time evolution (VarQITE) approach. Guided by McLachlan's variational principle, the VarQITE method provides analytical metrics and gradients, resulting in improved convergence efficiency and robustness of the QCMPS. We validate these improvements numerically through simulations of $\rm H_2$, $\rm H_4$, and $\rm LiH$ molecules. Additionally, given that VarQITE is applicable to non-Hermitian Hamiltonians, we evaluate its effectiveness in preparing the ground state of transcorrelated (TC) Hamiltonians. This approach yields energy estimates comparable to the complete basis set (CBS) limit while using even fewer qubits. Specifically, we perform simulations of the beryllium atom and $\rm LiH$ molecule using only three qubits, maintaining high fidelity with the CBS ground state energy of these systems. This qubit reduction is achieved through the combined advantages of both the QCMPS ansatz and transcorrelation. Our findings demonstrate the potential practicality of this quantum chemistry algorithm on near-term quantum devices., Comment: 15 pages, 8 figures

Existing methods for graph out-of-distribution (OOD) generalization primarily rely on empirical studies on synthetic datasets. Such approaches tend to overemphasize the causal relationships between invariant sub-graphs and labels, thereby neglecting the non-negligible role of environment in real-world scenarios. In contrast to previous studies that impose rigid independence assumptions on environments and invariant sub-graphs, this paper presents the theorems of environment-label dependency and mutable rationale invariance, where the former characterizes the usefulness of environments in determining graph labels while the latter refers to the mutable importance of graph rationales. Based on analytic investigations, a novel variational inference based method named ``Probability Dependency on Environments and Rationales for OOD Graphs on Real-world Data'' (DEROG) is introduced. To alleviate the adverse effect of unknown prior knowledge on environments and rationales, DEROG utilizes generalized Bayesian inference. Further, DEROG employs an EM-based algorithm for optimization. Finally, extensive experiments on real-world datasets under different distribution shifts are conducted to show the superiority of DEROG. Our code is publicly available at https://anonymous.4open.science/r/DEROG-536B., Comment: 21 pages, 5 figures

Generative models hold promise for revolutionizing medical education, robot-assisted surgery, and data augmentation for medical AI development. Diffusion models can now generate realistic images from text prompts, while recent advancements have demonstrated their ability to create diverse, high-quality videos. However, these models often struggle with generating accurate representations of medical procedures and detailed anatomical structures. This paper introduces Bora, the first spatio-temporal diffusion probabilistic model designed for text-guided biomedical video generation. Bora leverages Transformer architecture and is pre-trained on general-purpose video generation tasks. It is fine-tuned through model alignment and instruction tuning using a newly established medical video corpus, which includes paired text-video data from various biomedical fields. To the best of our knowledge, this is the first attempt to establish such a comprehensive annotated biomedical video dataset. Bora is capable of generating high-quality video data across four distinct biomedical domains, adhering to medical expert standards and demonstrating consistency and diversity. This generalist video generative model holds significant potential for enhancing medical consultation and decision-making, particularly in resource-limited settings. Additionally, Bora could pave the way for immersive medical training and procedure planning. Extensive experiments on distinct medical modalities such as endoscopy, ultrasound, MRI, and cell tracking validate the effectiveness of our model in understanding biomedical instructions and its superior performance across subjects compared to state-of-the-art generation models.

Evaluation is critical for assessing capabilities, tracking scientific progress, and informing model selection. In this paper, we present three desiderata for a good benchmark for language models: (i) salience (e.g., knowledge about World War II is more salient than a random day in history), (ii) novelty (i.e., the benchmark reveals new trends in model rankings not shown by previous benchmarks), and (iii) difficulty (i.e., the benchmark should be difficult for existing models, leaving headroom for future improvement). We operationalize these three desiderata and cast benchmark creation as a search problem, that of finding benchmarks that that satisfy all three desiderata. To tackle this search problem, we present AutoBencher, which uses a language model to automatically search for datasets that meet the three desiderata. AutoBencher uses privileged information (e.g. relevant documents) to construct reliable datasets, and adaptivity with reranking to optimize for the search objective. We use AutoBencher to create datasets for math, multilingual, and knowledge-intensive question answering. The scalability of AutoBencher allows it to test fine-grained categories and tail knowledge, creating datasets that are on average 27% more novel and 22% more difficult than existing benchmarks. A closer investigation of our constructed datasets shows that we can identify specific gaps in LM knowledge in language models that are not captured by existing benchmarks, such as Gemini Pro performing much worse on question answering about the Permian Extinction and Fordism, while OpenAGI-7B performing surprisingly well on QA about COVID-19., Comment: preprint

In recent years, the rapid increase in scientific papers has overwhelmed traditional review mechanisms, resulting in varying quality of publications. Although existing methods have explored the capabilities of Large Language Models (LLMs) for automated scientific reviewing, their generated contents are often generic or partial. To address the issues above, we introduce an automated paper reviewing framework SEA. It comprises of three modules: Standardization, Evaluation, and Analysis, which are represented by models SEA-S, SEA-E, and SEA-A, respectively. Initially, SEA-S distills data standardization capabilities of GPT-4 for integrating multiple reviews for a paper. Then, SEA-E utilizes standardized data for fine-tuning, enabling it to generate constructive reviews. Finally, SEA-A introduces a new evaluation metric called mismatch score to assess the consistency between paper contents and reviews. Moreover, we design a self-correction strategy to enhance the consistency. Extensive experimental results on datasets collected from eight venues show that SEA can generate valuable insights for authors to improve their papers., Comment: Accepted by EMNLP 2024

Cryogenic Electron Tomography (CryoET) is a useful imaging technology in structural biology that is hindered by its need for manual annotations, especially in particle picking. Recent works have endeavored to remedy this issue with few-shot learning or contrastive learning techniques. However, supervised training is still inevitable for them. We instead choose to leverage the power of existing 2D foundation models and present a novel, training-free framework, CryoSAM. In addition to prompt-based single-particle instance segmentation, our approach can automatically search for similar features, facilitating full tomogram semantic segmentation with only one prompt. CryoSAM is composed of two major parts: 1) a prompt-based 3D segmentation system that uses prompts to complete single-particle instance segmentation recursively with Cross-Plane Self-Prompting, and 2) a Hierarchical Feature Matching mechanism that efficiently matches relevant features with extracted tomogram features. They collaborate to enable the segmentation of all particles of one category with just one particle-specific prompt. Our experiments show that CryoSAM outperforms existing works by a significant margin and requires even fewer annotations in particle picking. Further visualizations demonstrate its ability when dealing with full tomogram segmentation for various subcellular structures. Our code is available at: https://github.com/xulabs/aitom, Comment: This preprint has not undergone peer review or any post-submission improvements or corrections. The Version of Record of this contribution will be published in MICCAI 2024

Understanding how light modifies long-range order is key to improve our ability to control material functionality on an ultrafast timescale. Transient spatial heterogeneity has been proposed in many materials, but isolating the dynamics of different regions experimentally has been challenging. Here we address this issue and measure the dynamics of orbital order melting in the layered manganite, La0.5Sr1.5MnO4, and isolate the surface dynamics from the bulk for the first time. Bulk measurements show orbital order is rapidly suppressed, but the correlation length surprisingly increases. However, the surface dynamics, show a stronger suppression and a significant decrease in correlation length. By isolating the surface changes, we find that light preferentially melts a less ordered surface and the loss of long-range order is likely driven by the formation of local and disordered polarons. Melting the disordered surface effectively increases the average correlation of the bulk probed volume, resolving the contradictory response. These results show that surface scattering methods are necessary to understand both surface and bulk dynamics in heterogeneous materials., Comment: 22 pages, 8 figures

Text-to-video (T2V) generation has recently garnered significant attention thanks to the large multi-modality model Sora. However, T2V generation still faces two important challenges: 1) Lacking a precise open sourced high-quality dataset. The previous popular video datasets, e.g. WebVid-10M and Panda-70M, are either with low quality or too large for most research institutions. Therefore, it is challenging but crucial to collect a precise high-quality text-video pairs for T2V generation. 2) Ignoring to fully utilize textual information. Recent T2V methods have focused on vision transformers, using a simple cross attention module for video generation, which falls short of thoroughly extracting semantic information from text prompt. To address these issues, we introduce OpenVid-1M, a precise high-quality dataset with expressive captions. This open-scenario dataset contains over 1 million text-video pairs, facilitating research on T2V generation. Furthermore, we curate 433K 1080p videos from OpenVid-1M to create OpenVidHD-0.4M, advancing high-definition video generation. Additionally, we propose a novel Multi-modal Video Diffusion Transformer (MVDiT) capable of mining both structure information from visual tokens and semantic information from text tokens. Extensive experiments and ablation studies verify the superiority of OpenVid-1M over previous datasets and the effectiveness of our MVDiT., Comment: 15 pages, 9 figures

We measure the performance of in-context learning as a function of task novelty and difficulty for open and closed questions. For that purpose, we created a novel benchmark consisting of hard scientific questions, each paired with a context of various relevancy. We show that counter-intuitively, a context that is more aligned with the topic does not always help more than a less relevant context. This effect is especially visible for open questions and questions of high difficulty or novelty. This result reveals a fundamental difference between the treatment of close-form and open-form questions by large-language models and shows a need for a more robust evaluation of in-context learning on the variety of different types of questions. It also poses a new question of how to optimally select a context for large language models, especially in the context of Retrieval Augmented Generation (RAG) systems. Our results suggest that the answer to this question can be highly application-dependent and might be contingent on factors including the format of the question, the perceived difficulty level of the questions, and the novelty or popularity of the information we seek., Comment: 8 pages plus references, 4 main figures, 6 pages of supplementary material

Large Language Models (LLMs) represent a significant stride toward Artificial General Intelligence. As scaling laws underscore the potential of increasing model sizes, the academic community has intensified its investigations into LLMs with capacities exceeding 50 billion parameters. This technical report builds on our prior work with Tele-FLM (also known as FLM-2), a publicly available 52-billion-parameter model. We delve into two primary areas: we first discuss our observation of Supervised Fine-tuning (SFT) on Tele-FLM-52B, which supports the "less is more" approach for SFT data construction; second, we demonstrate our experiments and analyses on the best practices for progressively growing a model from 52 billion to 102 billion, and subsequently to 1 trillion parameters. We will open-source a 1T model checkpoint, namely Tele-FLM-1T, to advance further training and research., Comment: For the Tele-FLM-52B tech report, see also 2404.16645

Recommendation systems, which assist users in discovering their preferred items among numerous options, have served billions of users across various online platforms. Intuitively, users' interactions with items are highly driven by their unchanging inherent intents (e.g., always preferring high-quality items) and changing demand intents (e.g., wanting a T-shirt in summer but a down jacket in winter). However, both types of intents are implicitly expressed in recommendation scenario, posing challenges in leveraging them for accurate intent-aware recommendations. Fortunately, in search scenario, often found alongside recommendation on the same online platform, users express their demand intents explicitly through their query words. Intuitively, in both scenarios, a user shares the same inherent intent and the interactions may be influenced by the same demand intent. It is therefore feasible to utilize the interaction data from both scenarios to reinforce the dual intents for joint intent-aware modeling. But the joint modeling should deal with two problems: 1) accurately modeling users' implicit demand intents in recommendation; 2) modeling the relation between the dual intents and the interactive items. To address these problems, we propose a novel model named Unified Dual-Intents Translation for joint modeling of Search and Recommendation (UDITSR). To accurately simulate users' demand intents in recommendation, we utilize real queries from search data as supervision information to guide its generation. To explicitly model the relation among the triplet , we propose a dual-intent translation propagation mechanism to learn the triplet in the same semantic space via embedding translations. Extensive experiments demonstrate that UDITSR outperforms SOTA baselines both in search and recommendation tasks.

Deep learning has deeply changed the paradigms of many research fields. At the heart of chemical and physical sciences is the accurate ab initio calculation of many-body wavefunction, which has become one of the most notable examples to demonstrate the power of deep learning in science. In particular, the introduction of deep learning into quantum Monte Carlo (QMC) has significantly advanced the frontier of ab initio calculation, offering a universal tool to solve the electronic structure of materials and molecules. Deep learning QMC architectures were initial designed and tested on small molecules, focusing on comparisons with other state-of-the-art ab initio methods. Methodological developments, including extensions to real solids and periodic models, have been rapidly progressing and reported applications are fast expanding. This review covers the theoretical foundation of deep learning QMC for solids, the neural network wavefunction ansatz, and various of other methodological developments. Applications on computing energy, electron density, electric polarization, force and stress of real solids are also reviewed. The methods have also been extended to other periodic systems and finite temperature calculations. The review highlights the potentials and existing challenges of deep learning QMC in materials chemistry and condensed matter physics.

LLMs with visual inputs, i.e., Vision Language Models (VLMs), have the capacity to process state information as visual-textual prompts and respond with policy decisions in text. We propose LLaRA: Large Language and Robotics Assistant, a framework that formulates robot action policy as conversations and provides improved action outputs when trained with auxiliary data that complements policy learning. We first introduce an automated pipeline to generate conversation-style instruction tuning data from existing behavior cloning data. Then we enrich the dataset in a self-supervised fashion by formulating six auxiliary tasks. A VLM finetuned with the resulting collection of datasets can generate meaningful robot action policy decisions. Our experiments across multiple simulated and real-world environments demonstrate the state-of-the-art performance of the proposed LLaRA framework. The code, datasets, and pretrained models are available at https://github.com/LostXine/LLaRA.

End-to-end neural diarization (EEND) models offer significant improvements over traditional embedding-based Speaker Diarization (SD) approaches but falls short on generalizing to long-form audio with large number of speakers. EEND-vector-clustering method mitigates this by combining local EEND with global clustering of speaker embeddings from local windows, but this requires an additional speaker embedding framework alongside the EEND module. In this paper, we propose a novel framework applying EEND both locally and globally for long-form audio without separate speaker embeddings. This approach achieves significant relative DER reduction of 13% and 10% over the conventional 1-pass EEND on Callhome American English and RT03-CTS datasets respectively and marginal improvements over EEND-vector-clustering without the need for additional speaker embeddings. Furthermore, we discuss the computational complexity of our proposed framework and explore strategies for reducing processing times., Comment: Accepted at INTERSPEECH 2024

Speaker Diarization (SD) systems are typically audio-based and operate independently of the ASR system in traditional speech transcription pipelines and can have speaker errors due to SD and/or ASR reconciliation, especially around speaker turns and regions of speech overlap. To reduce these errors, a Lexical Speaker Error Correction (LSEC), in which an external language model provides lexical information to correct the speaker errors, was recently proposed. Though the approach achieves good Word Diarization error rate (WDER) improvements, it does not use any additional acoustic information and is prone to miscorrections. In this paper, we propose to enhance and acoustically ground the LSEC system with speaker scores directly derived from the existing SD pipeline. This approach achieves significant relative WDER reductions in the range of 25-40% over the audio-based SD, ASR system and beats the LSEC system by 15-25% relative on RT03-CTS, Callhome American English and Fisher datasets., Comment: Accepted at INTERSPEECH 2024

Embodied agents require robust navigation systems to operate in unstructured environments, making the robustness of Simultaneous Localization and Mapping (SLAM) models critical to embodied agent autonomy. While real-world datasets are invaluable, simulation-based benchmarks offer a scalable approach for robustness evaluations. However, the creation of a challenging and controllable noisy world with diverse perturbations remains under-explored. To this end, we propose a novel, customizable pipeline for noisy data synthesis, aimed at assessing the resilience of multi-modal SLAM models against various perturbations. The pipeline comprises a comprehensive taxonomy of sensor and motion perturbations for embodied multi-modal (specifically RGB-D) sensing, categorized by their sources and propagation order, allowing for procedural composition. We also provide a toolbox for synthesizing these perturbations, enabling the transformation of clean environments into challenging noisy simulations. Utilizing the pipeline, we instantiate the large-scale Noisy-Replica benchmark, which includes diverse perturbation types, to evaluate the risk tolerance of existing advanced RGB-D SLAM models. Our extensive analysis uncovers the susceptibilities of both neural (NeRF and Gaussian Splatting -based) and non-neural SLAM models to disturbances, despite their demonstrated accuracy in standard benchmarks. Our code is publicly available at https://github.com/Xiaohao-Xu/SLAM-under-Perturbation., Comment: 50 pages. arXiv admin note: substantial text overlap with arXiv:2402.08125

Bi-level optimization (BO) has become a fundamental mathematical framework for addressing hierarchical machine learning problems. As deep learning models continue to grow in size, the demand for scalable bi-level optimization solutions has become increasingly critical. Traditional gradient-based bi-level optimization algorithms, due to their inherent characteristics, are ill-suited to meet the demands of large-scale applications. In this paper, we introduce $\textbf{F}$orward $\textbf{G}$radient $\textbf{U}$nrolling with $\textbf{F}$orward $\textbf{F}$radient, abbreviated as $(\textbf{FG})^2\textbf{U}$, which achieves an unbiased stochastic approximation of the meta gradient for bi-level optimization. $(\text{FG})^2\text{U}$ circumvents the memory and approximation issues associated with classical bi-level optimization approaches, and delivers significantly more accurate gradient estimates than existing large-scale bi-level optimization approaches. Additionally, $(\text{FG})^2\text{U}$ is inherently designed to support parallel computing, enabling it to effectively leverage large-scale distributed computing systems to achieve significant computational efficiency. In practice, $(\text{FG})^2\text{U}$ and other methods can be strategically placed at different stages of the training process to achieve a more cost-effective two-phase paradigm. Further, $(\text{FG})^2\text{U}$ is easy to implement within popular deep learning frameworks, and can be conveniently adapted to address more challenging zeroth-order bi-level optimization scenarios. We provide a thorough convergence analysis and a comprehensive practical discussion for $(\text{FG})^2\text{U}$, complemented by extensive empirical evaluations, showcasing its superior performance in diverse large-scale bi-level optimization tasks.

This paper proposes a novel framework to reinforce classification models using language-guided generated counterfactual images. Deep learning classification models are often trained using datasets that mirror real-world scenarios. In this training process, because learning is based solely on correlations with labels, there is a risk that models may learn spurious relationships, such as an overreliance on features not central to the subject, like background elements in images. However, due to the black-box nature of the decision-making process in deep learning models, identifying and addressing these vulnerabilities has been particularly challenging. We introduce a novel framework for reinforcing the classification models, which consists of a two-stage process. First, we identify model weaknesses by testing the model using the counterfactual image dataset, which is generated by perturbed image captions. Subsequently, we employ the counterfactual images as an augmented dataset to fine-tune and reinforce the classification model. Through extensive experiments on several classification models across various datasets, we revealed that fine-tuning with a small set of counterfactual images effectively strengthens the model., Comment: 6 pages, 4 figures

Generating free-text rationales is among the emergent capabilities of Large Language Models (LLMs). These rationales have been found to enhance LLM performance across various NLP tasks. Recently, there has been growing interest in using these rationales to provide insights for various important downstream tasks. In this paper, we analyze generated free-text rationales in tasks with subjective answers, emphasizing the importance of rationalization in such scenarios. We focus on pairwise argument ranking, a highly subjective task with significant potential for real-world applications, such as debate assistance. We evaluate the persuasiveness of rationales generated by nine LLMs to support their subjective choices. Our findings suggest that open-source LLMs, particularly Llama2-70B-chat, are capable of providing highly persuasive rationalizations, surpassing even GPT models. Additionally, our experiments show that rationale persuasiveness can be improved by controlling its parameters through prompting or through self-refinement.

We introduce a new benchmark designed to advance the development of general-purpose, large-scale vision-language models for remote sensing images. Although several vision-language datasets in remote sensing have been proposed to pursue this goal, existing datasets are typically tailored to single tasks, lack detailed object information, or suffer from inadequate quality control. Exploring these improvement opportunities, we present a Versatile vision-language Benchmark for Remote Sensing image understanding, termed VRSBench. This benchmark comprises 29,614 images, with 29,614 human-verified detailed captions, 52,472 object references, and 123,221 question-answer pairs. It facilitates the training and evaluation of vision-language models across a broad spectrum of remote sensing image understanding tasks. We further evaluated state-of-the-art models on this benchmark for three vision-language tasks: image captioning, visual grounding, and visual question answering. Our work aims to significantly contribute to the development of advanced vision-language models in the field of remote sensing. The data and code can be accessed at https://github.com/lx709/VRSBench., Comment: Submitted for consideration at a conference

Recent advancements in multimodal foundation models have showcased impressive capabilities in understanding and reasoning with visual and textual information. Adapting these foundation models trained for general usage to specialized domains like biomedicine requires large-scale domain-specific instruction datasets. While existing works have explored curating such datasets automatically, the resultant datasets are not explicitly aligned with domain expertise. In this work, we propose a data-centric framework, Biomedical Visual Instruction Tuning with Clinician Preference Alignment (BioMed-VITAL), that incorporates clinician preferences into both stages of generating and selecting instruction data for tuning biomedical multimodal foundation models. First, during the generation stage, we prompt the GPT-4V generator with a diverse set of clinician-selected demonstrations for preference-aligned data candidate generation. Then, during the selection phase, we train a separate selection model, which explicitly distills clinician and policy-guided model preferences into a rating function to select high-quality data for medical instruction tuning. Results show that the model tuned with the instruction-following data from our method demonstrates a significant improvement in open visual chat (18.5% relatively) and medical VQA (win rate up to 81.73%). Our instruction-following data and models are available at BioMed-VITAL.github.io.

We describe the design and performance of a magnetic bottle electron spectrometer~(MBES) for high-energy electron spectroscopy. Our design features a ${\sim2}$~m long electron drift tube and electrostatic retardation lens, achieving sub-electronvolt (eV) electron kinetic energy resolution for high energy (several hundred eV) electrons with close to 4$\pi$ collection efficiency. A segmented anode electron detector enables the simultaneous collection of photoelectron spectra in high resolution and high collection efficiency modes. This versatile instrument is installed at the TMO endstation at the LCLS x-ray free-electron laser (XFEL). In this paper, we demonstrate its high resolution, collection efficiency and spatial selectivity in measurements where it is coupled to an XFEL source. These combined characteristics are designed to enable high-resolution time-resolved measurements using x-ray photoelectron, absorption, and Auger-Meitner spectroscopy. We also describe the pervasive artifact in MBES time-of-flight spectra that arises from a periodic modulation in electron detection efficiency, and present a robust analysis procedure for its removal.

Tactile sensors, which provide information about the physical properties of objects, are an essential component of robotic systems. The visuotactile sensing technology with the merits of high resolution and low cost has facilitated the development of robotics from environment exploration to dexterous operation. Over the years, several reviews on visuotactile sensors for robots have been presented, but few of them discussed the significance of signal processing methods to visuotactile sensors. Apart from ingenious hardware design, the full potential of the sensory system toward designated tasks can only be released with the appropriate signal processing methods. Therefore, this paper provides a comprehensive review of visuotactile sensors from the perspective of signal processing methods and outlooks possible future research directions for visuotactile sensors., Comment: Accepted by IEEE Journal of Selected Topics in Signal Processing

Moir\'e superlattice designed in stacked van der Waals material provides a dynamic platform for hosting exotic and emergent condensed matter phenomena. However, the relevance of strong correlation effects and the large size of moir\'e unit cells pose significant challenges for traditional computational techniques. To overcome these challenges, we develop an unsupervised deep learning approach to uncover electronic phases emerging from moir\'e systems based on variational optimization of neural network many-body wavefunction. Our approach has identified diverse quantum states, including novel phases such as generalized Wigner crystals, Wigner molecular crystals, and previously unreported Wigner covalent crystals. These discoveries provide insights into recent experimental studies and suggest new phases for future exploration. They also highlight the crucial role of spin polarization in determining Wigner phases. More importantly, our proposed deep learning approach is proven general and efficient, offering a powerful framework for studying moir\'e physics.

The significance of intelligent sensing systems is growing in the realm of smart services. These systems extract relevant signal features and generate informative representations for particular tasks. However, building the feature extraction component for such systems requires extensive domain-specific expertise or data. The exceptionally rapid development of foundation models is likely to usher in newfound abilities in such intelligent sensing. We propose a new scheme for sensing model, which we refer to as semi-generalist sensing model (SGSM). SGSM is able to semiautomatically solve various tasks using relatively less task-specific labeled data compared to traditional systems. Built through the analysis of the common theoretical model, SGSM can depict different modalities, such as the acoustic and Wi-Fi signal. Experimental results on such two heterogeneous sensors illustrate that SGSM functions across a wide range of scenarios, thereby establishing its broad applicability. In some cases, SGSM even achieves better performance than sensor-specific specialized solutions. Wi-Fi evaluations indicate a 20\% accuracy improvement when applying SGSM to an existing sensing model.

Fine-tuning large language models (LLMs) for specific domain tasks has achieved great success in Text-to-SQL tasks. However, these fine-tuned models often face challenges with multi-turn Text-to-SQL tasks caused by ambiguous or unanswerable questions. It is desired to enhance LLMs to handle multiple types of questions in multi-turn Text-to-SQL tasks. To address this, we propose a novel data augmentation method, called QDA-SQL, which generates multiple types of multi-turn Q\&A pairs by using LLMs. In QDA-SQL, we introduce a novel data augmentation method incorporating validation and correction mechanisms to handle complex multi-turn Text-to-SQL tasks. Experimental results demonstrate that QDA-SQL enables fine-tuned models to exhibit higher performance on SQL statement accuracy and enhances their ability to handle complex, unanswerable questions in multi-turn Text-to-SQL tasks. The generation script and test set are released at https://github.com/mcxiaoxiao/QDA-SQL., Comment: 13 pages, 7 figures

Conditional visual generation has witnessed remarkable progress with the advent of diffusion models (DMs), especially in tasks like control-to-image generation. However, challenges such as expensive computational cost, high inference latency, and difficulties of integration with large language models (LLMs) have necessitated exploring alternatives to DMs. This paper introduces ControlVAR, a novel framework that explores pixel-level controls in visual autoregressive (VAR) modeling for flexible and efficient conditional generation. In contrast to traditional conditional models that learn the conditional distribution, ControlVAR jointly models the distribution of image and pixel-level conditions during training and imposes conditional controls during testing. To enhance the joint modeling, we adopt the next-scale AR prediction paradigm and unify control and image representations. A teacher-forcing guidance strategy is proposed to further facilitate controllable generation with joint modeling. Extensive experiments demonstrate the superior efficacy and flexibility of ControlVAR across various conditional generation tasks against popular conditional DMs, \eg, ControlNet and T2I-Adaptor. Code: \url{https://github.com/lxa9867/ControlVAR}., Comment: 25 pages, 19 figures, 4 tables

The Large Language models (LLMs) have demonstrated supreme capabilities in text understanding and generation, but cannot be directly applied to cross-modal tasks without fine-tuning. This paper proposes a cross-modal in-context learning approach, empowering the frozen LLMs to achieve multiple audio tasks in a few-shot style without any parameter update. Specifically, we propose a novel and LLMs-driven audio codec model, LLM-Codec, to transfer the audio modality into the textual space, \textit{i.e.} representing audio tokens with words or sub-words in the vocabulary of LLMs, while keeping high audio reconstruction quality. The key idea is to reduce the modality heterogeneity between text and audio by compressing the audio modality into a well-trained LLMs token space. Thus, the audio representation can be viewed as a new \textit{foreign language}, and LLMs can learn the new \textit{foreign language} with several demonstrations. In experiments, we investigate the performance of the proposed approach across multiple audio understanding and generation tasks, \textit{e.g.} speech emotion classification, audio classification, text-to-speech generation, speech enhancement, etc. The experimental results demonstrate that the LLMs equipped with the proposed LLM-Codec, named as UniAudio 1.5, prompted by only a few examples, can achieve the expected functions in simple scenarios. It validates the feasibility and effectiveness of the proposed cross-modal in-context learning approach. To facilitate research on few-shot audio task learning and multi-modal LLMs, we have open-sourced the LLM-Codec model.

The fallback disk model is widely accepted to explain long-period neutron stars (NSs) which can't be simulated by magnetic dipole radiation. However, no confirmed detection of disk was found from the newly discovered long period pulsars GLEAM-X 162759.5-523504.3, GPM J1839-10 and the known slowest isolated NSs 1E 161348-5055. This might be that the disks have either been in noninteracting/inactive state where its emission is too weak to be detected or have been disrupted. In this work, we conduct simulations to examine the lifetime of supernova fallback disks around isolated neutron stars. We assume that the disk's mass varies in a self-similar way and its interaction with the NS occurs only in interacting/active state. Our results reveal that nearly all the interacting lifetimes for the disk are shorter than 0.1 Myr while the existence lifetimes are considerably longer., Comment: Accepted in ApJ, comments are welcome

Network dismantling aims to maximize the disintegration of a network by removing a specific set of nodes or edges and is applied to various tasks in diverse domains, such as cracking down on crime organizations, delaying the propagation of rumors, and blocking the transmission of viruses. Most of the current network dismantling methods are tailored for unsigned networks, which only consider the connection between nodes without evaluating the nature of the relationships, such as friendship/hostility, enhancing/repressing, and trust/distrust. We here propose an embedding-based algorithm, namely ESND, to solve the signed network dismantling problem. The algorithm generally iterates the following four steps, i.e., giant component detection, network embedding, node clustering, and removal node selection. To illustrate the efficacy and stability of ESND, we conduct extensive experiments on six signed network datasets as well as null models, and compare the performance of our method with baselines. Experimental results consistently show that the proposed ESND is superior to the baselines and displays stable performance with the change in the network structure. Additionally, we examine the impact of sign proportions on network robustness via ESND, observing that networks with a high ratio of negative edges are generally easier to dismantle than networks with high positive edges.

Learning-based approaches have witnessed great successes in blind single image super-resolution (SISR) tasks, however, handcrafted kernel priors and learning based kernel priors are typically required. In this paper, we propose a Meta-learning and Markov Chain Monte Carlo (MCMC) based SISR approach to learn kernel priors from organized randomness. In concrete, a lightweight network is adopted as kernel generator, and is optimized via learning from the MCMC simulation on random Gaussian distributions. This procedure provides an approximation for the rational blur kernel, and introduces a network-level Langevin dynamics into SISR optimization processes, which contributes to preventing bad local optimal solutions for kernel estimation. Meanwhile, a meta-learning-based alternating optimization procedure is proposed to optimize the kernel generator and image restorer, respectively. In contrast to the conventional alternating minimization strategy, a meta-learning-based framework is applied to learn an adaptive optimization strategy, which is less-greedy and results in better convergence performance. These two procedures are iteratively processed in a plug-and-play fashion, for the first time, realizing a learning-based but plug-and-play blind SISR solution in unsupervised inference. Extensive simulations demonstrate the superior performance and generalization ability of the proposed approach when comparing with state-of-the-arts on synthesis and real-world datasets. The code is available at https://github.com/XYLGroup/MLMC., Comment: This paper has been accepted for publication in IEEE Transactions on Pattern Analysis and Machine Intelligence (2024)

We elaborate on a new recursive method proposed in arXiv:2312.15484 for computing tree-level $n$-point supergluon amplitudes as well as those with one gluon, i.e., spinning amplitudes, in ${\rm AdS}_5 \times S^3$. We present an improved proof for the so-called "constructibility" of supergluon and spinning amplitudes based on their factorizations and flat-space limit, which allows us to determine these amplitudes in Mellin space to all $n$. We present explicit and remarkably simple expressions for up to $n=7$ supergluon amplitudes and $n=6$ spinning amplitudes, which can be viewed as AdS generalizations of the scalar-scaffolded gluon amplitudes proposed recently. We then reveal a series of hidden structures of these AdS amplitudes including (1) an understanding of general pole structures especially the precise truncation on descendent poles (2) a derivation of simple "Feynman rules" for the all-$n$ amplitudes with the simplest R-symmetry structures, and (3) certain universal behavior analogous to the soft/collinear limit of flat-space amplitudes., Comment: 46 pages, multiple figures

In reality, data often exhibit associations with multiple labels, making multi-label learning (MLL) become a prominent research topic. The last two decades have witnessed the success of MLL, which is indispensable from complete and accurate supervised information. However, obtaining such information in practice is always laborious and sometimes even impossible. To circumvent this dilemma, incomplete multi-label learning (InMLL) has emerged, aiming to learn from incomplete labeled data. To date, enormous InMLL works have been proposed to narrow the performance gap with complete MLL, whereas a systematic review for InMLL is still absent. In this paper, we not only attempt to fill the lacuna but also strive to pave the way for innovative research. Specifically, we retrospect the origin of InMLL, analyze the challenges of InMLL, and make a taxonomy of InMLL from the data-oriented and algorithm-oriented perspectives, respectively. Besides, we also present real applications of InMLL in various domains. More importantly, we highlight several potential future trends, including four open problems that are more in line with practice and three under-explored/unexplored techniques in addressing the challenges of InMLL, which may shed new light on developing novel research directions in the field of InMLL., Comment: 10 pages, 3 figures

The emergence of Large Language Models (LLMs) has necessitated the adoption of parallel training techniques, involving the deployment of thousands of GPUs to train a single model. Unfortunately, we have found that the efficiency of current parallel training is often suboptimal, largely due to the following two main issues. Firstly, hardware failures are inevitable, leading to interruptions in the training tasks. The inability to quickly identify the faulty components results in a substantial waste of GPU resources. Secondly, since GPUs must wait for parameter synchronization to complete before proceeding to the next round of computation, network congestions can greatly increase the waiting time for GPUs. To address these challenges, this paper introduces a communication-driven solution, namely the C4. The key insights of C4 are two folds. First, in parallel training, collective communication exhibits periodic and homogeneous characteristics, so any anomalies are certainly due to some form of hardware malfunction. By leveraging this feature, C4 can rapidly identify the faulty components, swiftly isolate the anomaly, and restart the task, thereby avoiding resource wastage caused by delays in anomaly detection. Second, the predictable communication model of collective communication, involving few large flows, allows C4 to efficiently execute traffic planning, substantially reducing network congestion. C4 has been extensively implemented across our production systems, cutting error-induced overhead by roughly 30% and enhancing runtime performance by about 15% for certain applications with moderate communication costs.

Large language models have demonstrated impressive performance on commonsense tasks; however, these tasks are often posed as multiple-choice questions, allowing models to exploit systematic biases. Commonsense is also inherently probabilistic with multiple correct answers. The purpose of "boiling water" could be making tea and cooking, but it also could be killing germs. Existing tasks do not capture the probabilistic nature of common sense. To this end, we present commonsense frame completion (CFC), a new generative task that evaluates common sense via multiple open-ended generations. We also propose a method of probabilistic evaluation that strongly correlates with human judgments. Humans drastically outperform strong language model baselines on our dataset, indicating this approach is both a challenging and useful evaluation of machine common sense., Comment: ACL 2024 Camera Ready