Search

In anomaly detection, methods based on large language models (LLMs) can incorporate expert knowledge, while task-specific smaller models excel at extracting normal patterns and detecting value fluctuations. Inspired by the human nervous system, where the brain stores expert knowledge and the peripheral nervous system and spinal cord handle specific tasks like withdrawal and knee-jerk reflexes, we propose CoLLaTe, a framework designed to facilitate collaboration between LLMs and task-specific models, leveraging the strengths of both. In this work, we first formulate the collaboration process and identify two key challenges in the collaboration between LLMs and task-specific models: (1) the misalignment between the expression domains of LLMs and smaller models, and (2) error accumulation arising from the predictions of both models. To address these challenges, we introduce two key components in CoLLaTe: the alignment module and the collaborative loss function. Through theoretical analysis and experimental validation, we demonstrate that these components effectively mitigate the identified challenges and achieve better performance than LLM based methods and task-specific smaller model.

Spectrally selective 3D chemical exchange saturation transfer (CEST) imaging typically demands single-shot readouts for the sake of time-saving, so as to simultaneously fit multiple CEST effects. However, up to now, only fast low angle shot (FLASH) and echo planar imaging (EPI) have been predominantly utilized for this purpose. In this work, we developed a 3D single-shot CEST sequence using true fast imaging with steady-state precession (True FISP) readout and optimized the scanning parameters through simulations. The performance of the CEST sequence was validated using an egg white phantom, ten healthy volunteers, and a patient with a brain tumor on a 3T human scanner. Subsequently, the proposed CEST sequence utilizing True FISP was compared to the commonly used FLASH-based CEST sequence. This comparison focused on SNR and image contrast, maintaining identical pre-saturation modes, TR, TE and scanning times. In the simulation experiments, the maximum CEST signal obtained from the True FISP was significantly greater than that obtained from the FLASH sequence. In the egg white phantom, the SNRs of amide proton transfer (APT) and nuclear Overhauser enhancement (NOE) effect images obtained from the True FISP were 68.3% and 57.0% higher than those obtained from the FLASH sequence, respectively. In healthy volunteers, saturated images collected with the True FISP sequence at 3.5 ppm showed an approximate 84% increase in mean temporal SNR compared to those collected with the FLASH sequence. Compared to the FLASH sequence, the CEST images obtained from the True FISP sequence could display more detailed brain tissue structures of both normal individuals and the patient with a brain tumor. Therefore, due to the high SNR inherent in the sequence, True FISP has the potential to be used for fast and high-quality 3D image readout of CEST contrasts in clinical applications., Comment: 37 pages, 8 figures

In this work, we study the experts problem in the distributed setting where an expert's cost needs to be aggregated across multiple servers. Our study considers various communication models such as the message-passing model and the broadcast model, along with multiple aggregation functions, such as summing and taking the $\ell_p$ norm of an expert's cost across servers. We propose the first communication-efficient protocols that achieve near-optimal regret in these settings, even against a strong adversary who can choose the inputs adaptively. Additionally, we give a conditional lower bound showing that the communication of our protocols is nearly optimal. Finally, we implement our protocols and demonstrate empirical savings on the HPO-B benchmarks., Comment: 38th Conference on Neural Information Processing Systems (NeurIPS 2024)

Satisfactory progress has been achieved recently in universal segmentation of CT images. Following the success of vision-language methods, there is a growing trend towards utilizing text prompts and contrastive learning to develop universal segmentation models. However, there exists a significant imbalance in information density between 3D images and text prompts. Moreover, the standard fully connected layer segmentation approach faces significant challenges in handling multiple classes and exhibits poor generalizability. To address these challenges, we propose the VOxel Interacting with LAnguage method (VOILA) for universal CT image segmentation. Initially, we align voxels and language into a shared representation space and classify voxels on the basis of cosine similarity. Subsequently, we develop the Voxel-Language Interaction framework to mitigate the impact of class imbalance caused by foreground-background discrepancies and variations in target volumes. Furthermore, a Complexity-Aware Sampling method is proposed to focus on region hard to segment, achieved by generating pseudo-heatmaps from a trainable Gaussian mixture distribution. Our results indicate the proposed VOILA is capable to achieve improved performance with reduced parameters and computational cost during training. Furthermore, it demonstrates significant generalizability across diverse datasets without additional fine-tuning., Comment: Accepted by AAAI 2025

With the rapid scaling of large language models (LLMs), structured pruning has become a widely used technique to learn efficient, smaller models from larger ones, delivering superior performance compared to training similarly sized models from scratch. In this paper, we move beyond the traditional static pruning approach of determining a fixed pruning mask for a model, and propose a dynamic approach to structured pruning. In our method, the pruning mask is input-dependent and adapts dynamically based on the information described in a user instruction. Our approach, termed "instruction-following pruning", introduces a sparse mask predictor that takes the user instruction as input and dynamically selects the most relevant model parameters for the given task. To identify and activate effective parameters, we jointly optimize the sparse mask predictor and the LLM, leveraging both instruction-following data and the pre-training corpus. Experimental results demonstrate the effectiveness of our approach on a wide range of evaluation benchmarks. For example, our 3B activated model improves over the 3B dense model by 5-8 points of absolute margin on domains such as math and coding, and rivals the performance of a 9B model., Comment: 13 pages, 3 figures

Continuous gravitational waves (CWs) emission from neutron stars carries information about their internal structure and equation of state, and it can provide tests of General Relativity. We present a search for CWs from a set of 45 known pulsars in the first part of the fourth LIGO--Virgo--KAGRA observing run, known as O4a. We conducted a targeted search for each pulsar using three independent analysis methods considering the single-harmonic and the dual-harmonic emission models. We find no evidence of a CW signal in O4a data for both models and set upper limits on the signal amplitude and on the ellipticity, which quantifies the asymmetry in the neutron star mass distribution. For the single-harmonic emission model, 29 targets have the upper limit on the amplitude below the theoretical spin-down limit. The lowest upper limit on the amplitude is $6.4\!\times\!10^{-27}$ for the young energetic pulsar J0537-6910, while the lowest constraint on the ellipticity is $8.8\!\times\!10^{-9}$ for the bright nearby millisecond pulsar J0437-4715. Additionally, for a subset of 16 targets we performed a narrowband search that is more robust regarding the emission model, with no evidence of a signal. We also found no evidence of non-standard polarizations as predicted by the Brans-Dicke theory., Comment: main paper: 12 pages, 6 figures, 4 tables

Measuring inequality of opportunities has long been a challenging and open problem, primarily due to the limitations associated with individual-level data. In this study, we utilize data obtained from vehicle license plates in a comprehensive survey (17258 vehicles from 6 major cities in China) to evaluate the inequality of opportunities in the country. In our context, we define inequality of opportunity as the scenario where relatively expensive vehicles have a higher likelihood of being paired with license plates featuring 'Lucky Numbers'. To quantify this, we propose a lucky-number-based opportunity Gini coefficient. Through the calculation of the opportunity Gini coefficient, we observe a significant and positive correlation between opportunity inequality and income inequality. Particularly noteworthy is our finding that the advancement of technology, exemplified by the widespread adoption of new energy vehicles, can substantially reduce the inequality of opportunity. Taking incorporation of a random lottery process before acquiring a motor vehicle in Beijing and Shanghai as a natural experiment, our empirical results support the argument that, in terms of equality, employing random drawing is a fair and equitable approach for allocating scarce resources., Comment: Licence plate numbers in other major cities have not been covered (Guangzhou, Shenzhen), and hence new data still needs to be supplemented. Single GINI coefficient is insufficient to support conclusions in terms of inequality of opportunity, and we are still processing other methods. Therefore, this work is incomplete, and has been withdrawn by authors

In the face of escalating surveillance and censorship within the cyberspace, the sanctity of personal privacy has come under siege, necessitating the development of steganography, which offers a way to securely hide messages within innocent-looking texts. Previous methods alternate the texts to hide private massages, which is not secure. Large Language Models (LLMs) provide high-quality and explicit distribution, which is an available mathematical tool for secure steganography methods. However, existing attempts fail to achieve high capacity, time efficiency and correctness simultaneously, and their strongly coupling designs leave little room for refining them to achieve better performance. To provide a secure, high-capacity and efficient steganography method, we introduce ShiMer. Specifically, ShiMer pseudorandomly shifts the probability interval of the LLM's distribution to obtain a private distribution, and samples a token according to the private bits. ShiMer produced steganographic texts are indistinguishable in quality from the normal texts directly generated by the language model. To further enhance the capacity of ShiMer, we design a reordering algorithm to minimize the occurrence of interval splitting during decoding phase. Experimental results indicate that our method achieves the highest capacity and efficiency among existing secure steganography techniques.

We present OLMo 2, the next generation of our fully open language models. OLMo 2 includes dense autoregressive models with improved architecture and training recipe, pretraining data mixtures, and instruction tuning recipes. Our modified model architecture and training recipe achieve both better training stability and improved per-token efficiency. Our updated pretraining data mixture introduces a new, specialized data mix called Dolmino Mix 1124, which significantly improves model capabilities across many downstream task benchmarks when introduced via late-stage curriculum training (i.e. specialized data during the annealing phase of pretraining). Finally, we incorporate best practices from T\"ulu 3 to develop OLMo 2-Instruct, focusing on permissive data and extending our final-stage reinforcement learning with verifiable rewards (RLVR). Our OLMo 2 base models sit at the Pareto frontier of performance to compute, often matching or outperforming open-weight only models like Llama 3.1 and Qwen 2.5 while using fewer FLOPs and with fully transparent training data, code, and recipe. Our fully open OLMo 2-Instruct models are competitive with or surpassing open-weight only models of comparable size, including Qwen 2.5, Llama 3.1 and Gemma 2. We release all OLMo 2 artifacts openly -- models at 7B and 13B scales, both pretrained and post-trained, including their full training data, training code and recipes, training logs and thousands of intermediate checkpoints. The final instruction model is available on the Ai2 Playground as a free research demo., Comment: Model demo available at playground.allenai.org

Visual Instruction Tuning typically requires a large amount of vision-language training data. This data often containing redundant information that increases computational costs without proportional performance gains. In this work, we introduce ICONS, a gradient-driven Influence CONsensus approach for vision-language data Selection that selects a compact training dataset for efficient multi-task training. The key element of our approach is cross-task influence consensus, which uses majority voting across task-specific influence matrices to identify samples that are consistently valuable across multiple tasks, allowing us to effectively prioritize data that optimizes for overall performance. Experiments show that models trained on our selected data (20% of LLaVA-665K) achieve 98.6% of the relative performance obtained using the full dataset. Additionally, we release this subset, LLaVA-ICONS-133K, a compact yet highly informative subset of LLaVA-665K visual instruction tuning data, preserving high impact training data for efficient vision-language model development., Comment: 25 pages, 19 figures

This study presents a machine learning approach to predict the number of barges transported by vessels on inland waterways using tracking data from the Automatic Identification System (AIS). While AIS tracks the location of tug and tow vessels, it does not monitor the presence or number of barges transported by those vessels. Understanding the number and types of barges conveyed along river segments, between ports, and at ports is crucial for estimating the quantities of freight transported on the nation's waterways. This insight is also valuable for waterway management and infrastructure operations impacting areas such as targeted dredging operations, and data-driven resource allocation. Labeled sample data was generated using observations from traffic cameras located along key river segments and matched to AIS data records. A sample of 164 vessels representing up to 42 barge convoys per vessel was used for model development. The methodology involved first predicting barge presence and then predicting barge quantity. Features derived from the AIS data included speed measures, vessel characteristics, turning measures, and interaction terms. For predicting barge presence, the AdaBoost model achieved an F1 score of 0.932. For predicting barge quantity, the Random Forest combined with an AdaBoost ensemble model achieved an F1 score of 0.886. Bayesian optimization was used for hyperparameter tuning. By advancing predictive modeling for inland waterways, this study offers valuable insights for transportation planners and organizations, which require detailed knowledge of traffic volumes, including the flow of commodities, their destinations, and the tonnage moving in and out of ports.

The remarkable performance of models like the OpenAI o1 can be attributed to their ability to emulate human-like long-time thinking during inference. These models employ extended chain-of-thought (CoT) processes, exploring multiple strategies to enhance problem-solving capabilities. However, a critical question remains: How to intelligently and efficiently scale computational resources during testing. This paper presents the first comprehensive study on the prevalent issue of overthinking in these models, where excessive computational resources are allocated for simple problems with minimal benefit. We introduce novel efficiency metrics from both outcome and process perspectives to evaluate the rational use of computational resources by o1-like models. Using a self-training paradigm, we propose strategies to mitigate overthinking, streamlining reasoning processes without compromising accuracy. Experimental results show that our approach successfully reduces computational overhead while preserving model performance across a range of testsets with varying difficulty levels, such as GSM8K, MATH500, GPQA, and AIME., Comment: Work in progress

Quantifying image complexity at the entity level is straightforward, but the assessment of semantic complexity has been largely overlooked. In fact, there are differences in semantic complexity across images. Images with richer semantics can tell vivid and engaging stories and offer a wide range of application scenarios. For example, the Cookie Theft picture is such a kind of image and is widely used to assess human language and cognitive abilities due to its higher semantic complexity. Additionally, semantically rich images can benefit the development of vision models, as images with limited semantics are becoming less challenging for them. However, such images are scarce, highlighting the need for a greater number of them. For instance, there is a need for more images like Cookie Theft to cater to people from different cultural backgrounds and eras. Assessing semantic complexity requires human experts and empirical evidence. Automatic evaluation of how semantically rich an image will be the first step of mining or generating more images with rich semantics, and benefit human cognitive assessment, Artificial Intelligence, and various other applications. In response, we propose the Image Semantic Assessment (ISA) task to address this problem. We introduce the first ISA dataset and a novel method that leverages language to solve this vision problem. Experiments on our dataset demonstrate the effectiveness of our approach. Our data and code are available at: https://github.com/xiujiesong/ISA., Comment: Accepted by AAAI 2025

We address protein structure prediction in the 3D Hydrophobic-Polar lattice model through two novel deep learning architectures. For proteins under 36 residues, our hybrid reservoir-based model combines fixed random projections with trainable deep layers, achieving optimal conformations with 25% fewer training episodes. For longer sequences, we employ a long short-term memory network with multi-headed attention, matching best-known energy values. Both architectures leverage a stabilized Deep Q-Learning framework with experience replay and target networks, demonstrating consistent achievement of optimal conformations while significantly improving training efficiency compared to existing methods., Comment: 15 pages, 9 figures

Archaeological catalogs, containing key elements such as artifact images, morphological descriptions, and excavation information, are essential for studying artifact evolution and cultural inheritance. These data are widely scattered across publications, requiring automated collection methods. However, existing Large Vision-Language Models (VLMs) and their derivative data collection methods face challenges in accurate image detection and modal matching when processing archaeological catalogs, making automated collection difficult. To address these issues, we propose a novel archaeological catalog collection method based on Large Vision-Language Models that follows an approach comprising three modules: document localization, block comprehension and block matching. Through practical data collection from the Dabagou and Miaozigou pottery catalogs and comparison experiments, we demonstrate the effectiveness of our approach, providing a reliable solution for automated collection of archaeological catalogs., Comment: 4 pages,4 figures,www source track

We present a novel constraint on light dark matter utilizing $1.54$ tonne$\cdot$year of data acquired from the PandaX-4T dual-phase xenon time projection chamber. This constraint is derived through detecting electronic recoil signals resulting from the interaction with solar-enhanced dark matter flux. Low-mass dark matter particles, lighter than a few MeV/$c^2$, can scatter with the thermal electrons in the Sun. Consequently, with higher kinetic energy, the boosted dark matter component becomes detectable via contact scattering with xenon electrons, resulting in a few keV energy deposition that exceeds the threshold of PandaX-4T. We calculate the expected recoil energy in PandaX-4T considering the Sun's acceleration and the detection capabilities of the xenon detector. The first experimental search results using the xenon detector yield the most stringent cross-section of $3.51 \times 10^{-39}~\mathrm{cm}^2$ at $0.08~\mathrm{MeV}$/$c^2$ for a solar boosted dark matter mass ranging from $0.02$ to $10~ \mathrm{MeV}$/$c^2$, achieving a 23 fold improvement compared with earlier experimental studies.

Linguistic steganography embeds secret information in seemingly innocent texts, safeguarding privacy in surveillance environments. Generative linguistic steganography leverages the probability distribution of language models (LMs) and applies steganographic algorithms to generate stego tokens, gaining attention with recent Large Language Model (LLM) advancements. To enhance security, researchers develop distribution-preserving stego algorithms to minimize the gap between stego sampling and LM sampling. However, the reliance on language model distributions, coupled with deviations from real-world cover texts, results in insufficient imperceptibility when facing steganalysis detectors in real-world scenarios. Moreover, LLM distributions tend to be more deterministic, resulting in reduced entropy and, consequently, lower embedding capacity. In this paper, we propose FreStega, a plug-and-play method to reconstruct the distribution of language models used for generative linguistic steganography. FreStega dynamically adjusts token probabilities from the language model at each step of stegotext auto-regressive generation, leveraging both sequential and spatial dimensions. In sequential adjustment, the temperature is dynamically adjusted based on instantaneous entropy, enhancing the diversity of stego texts and boosting embedding capacity. In the spatial dimension, the distribution is aligned with guidance from the target domain corpus, closely mimicking real cover text in the target domain. By reforming the distribution, FreStega enhances the imperceptibility of stego text in practical scenarios and improves steganographic capacity by 15.41\%, all without compromising the quality of the generated text. FreStega serves as a plug-and-play remedy to enhance the imperceptibility and embedding capacity of existing distribution-preserving steganography methods in real-world scenarios.

We show that the traditional moments approach in lattice QCD, based on operator product expansion (OPE), can be realized in a way that avoids power divergent mixings, and thus allows to extract moments order by order, to all orders in principle. We propose to use momentum instead of distance as a leverage to extract moments at a given order. Moreover, by utilizing symmetry of the lattice matrix elements, we can determine the even and odd moments separately. As a demonstrative example, we determine the first three moments beyond the tensor charge gT of the isovector quark transversity distribution in the nucleon., Comment: 5 pages, 1 figure

Camouflaged Object Detection (COD) is designed to identify objects that blend seamlessly with their surroundings. Due to the complexity of camouflaged objects (such as shape, color, and texture), their semantic cues are often blurred or completely lost, posing a significant challenge for COD. Existing COD methods often rely on visual features, which are not stable enough in changeable camouflage environments. This instability leads to false positives and false negatives, resulting in incomplete or inaccurate segmentation results. In this paper, to solve this problem, we propose a new task, Class-Guided Camouflaged Object Detection (CG-COD), which extends the traditional COD task by introducing object class knowledge, significantly improving the robustness and segmentation accuracy of the model in complex environments. Toward this end, we construct a dataset, CamoClass, containing the camouflaged objects in the real scenes and their corresponding class annotation. Based on this, we propose a multi-stage framework CGNet which consists of a plug-and-play class prompt generator and a class-guided detector. Under the guidance of textual information, CGNet enables efficient segmentation. It is worth emphasizing that for the first time, we extend the object class annotations on existing COD benchmark datasets, and introduce a flexible framework to improve the performance of the existing COD model under text guidance.

Orientation is a key attribute of objects, crucial for understanding their spatial pose and arrangement in images. However, practical solutions for accurate orientation estimation from a single image remain underexplored. In this work, we introduce Orient Anything, the first expert and foundational model designed to estimate object orientation in a single- and free-view image. Due to the scarcity of labeled data, we propose extracting knowledge from the 3D world. By developing a pipeline to annotate the front face of 3D objects and render images from random views, we collect 2M images with precise orientation annotations. To fully leverage the dataset, we design a robust training objective that models the 3D orientation as probability distributions of three angles and predicts the object orientation by fitting these distributions. Besides, we employ several strategies to improve synthetic-to-real transfer. Our model achieves state-of-the-art orientation estimation accuracy in both rendered and real images and exhibits impressive zero-shot ability in various scenarios. More importantly, our model enhances many applications, such as comprehension and generation of complex spatial concepts and 3D object pose adjustment., Comment: Project Page: https://orient-anything.github.io/

Deep reinforcement learning (DRL) shows promising potential for autonomous driving decision-making. However, DRL demands extensive computational resources to achieve a qualified policy in complex driving scenarios due to its low learning efficiency. Moreover, leveraging expert guidance from human to enhance DRL performance incurs prohibitively high labor costs, which limits its practical application. In this study, we propose a novel large language model (LLM) guided deep reinforcement learning (LGDRL) framework for addressing the decision-making problem of autonomous vehicles. Within this framework, an LLM-based driving expert is integrated into the DRL to provide intelligent guidance for the learning process of DRL. Subsequently, in order to efficiently utilize the guidance of the LLM expert to enhance the performance of DRL decision-making policies, the learning and interaction process of DRL is enhanced through an innovative expert policy constrained algorithm and a novel LLM-intervened interaction mechanism. Experimental results demonstrate that our method not only achieves superior driving performance with a 90\% task success rate but also significantly improves the learning efficiency and expert guidance utilization efficiency compared to state-of-the-art baseline algorithms. Moreover, the proposed method enables the DRL agent to maintain consistent and reliable performance in the absence of LLM expert guidance. The code and supplementary videos are available at https://bitmobility.github.io/LGDRL/.

In Continual Learning (CL), while existing work primarily focuses on the multi-class classification task, there has been limited research on Multi-Label Learning (MLL). In practice, MLL datasets are often class-imbalanced, making it inherently challenging, a problem that is even more acute in CL. Due to its sensitivity to imbalance, Macro-AUC is an appropriate and widely used measure in MLL. However, there is no research to optimize Macro-AUC in MLCL specifically. To fill this gap, in this paper, we propose a new memory replay-based method to tackle the imbalance issue for Macro-AUC-oriented MLCL. Specifically, inspired by recent theory work, we propose a new Reweighted Label-Distribution-Aware Margin (RLDAM) loss. Furthermore, to be compatible with the RLDAM loss, a new memory-updating strategy named Weight Retain Updating (WRU) is proposed to maintain the numbers of positive and negative instances of the original dataset in memory. Theoretically, we provide superior generalization analyses of the RLDAM-based algorithm in terms of Macro-AUC, separately in batch MLL and MLCL settings. This is the first work to offer theoretical generalization analyses in MLCL to our knowledge. Finally, a series of experimental results illustrate the effectiveness of our method over several baselines. Our codes are available at https://github.com/ML-Group-SDU/Macro-AUC-CL., Comment: 7 pages of main text, 11 pages of appendix, accepted to AAAI 2025

Time series forecasting has important applications across diverse domains. EasyTime, the system we demonstrate, facilitates easy use of time-series forecasting methods by researchers and practitioners alike. First, EasyTime enables one-click evaluation, enabling researchers to evaluate new forecasting methods using the suite of diverse time series datasets collected in the preexisting time series forecasting benchmark (TFB). This is achieved by leveraging TFB's flexible and consistent evaluation pipeline. Second, when practitioners must perform forecasting on a new dataset, a nontrivial first step is often to find an appropriate forecasting method. EasyTime provides an Automated Ensemble module that combines the promising forecasting methods to yield superior forecasting accuracy compared to individual methods. Third, EasyTime offers a natural language Q&A module leveraging large language models. Given a question like "Which method is best for long term forecasting on time series with strong seasonality?", EasyTime converts the question into SQL queries on the database of results obtained by TFB and then returns an answer in natural language and charts. By demonstrating EasyTime, we intend to show how it is possible to simplify the use of time series forecasting and to offer better support for the development of new generations of time series forecasting methods., Comment: Accepted by ICDE2025

The advanced bidirectional EV charging and discharging technology, aimed at supporting grid stability and emergency operations, has driven a growing interest in workplace applications. It not only effectively reduces electricity expenses but also enhances the resilience of handling practical issues, such as peak power limitation, fluctuating energy prices, and unpredictable EV departures. However, existing EV charging strategies have yet to fully consider these factors in a way that benefits both office buildings and EV users simultaneously. To address these issues, we propose HUCA, a novel real-time charging control for regulating energy demands for both the building and electric vehicles. HUCA employs hierarchical actor-critic networks to dynamically reduce electricity costs in buildings, accounting for the needs of EV charging in the dynamic pricing scenario. To tackle the uncertain EV departures, a new critic augmentation is introduced to account for departure uncertainties in evaluating the charging decisions, while maintaining the robustness of the charging control. Experiments on real-world electricity datasets under both simulated certain and uncertain departure scenarios demonstrate that HUCA outperforms baselines in terms of total electricity costs while maintaining competitive performance in fulfilling EV charging requirements. A case study also manifests that HUCA effectively balances energy supply between the building and EVs based on real-time information.

Aims. We reported the discovery that a changing-look AGN SDSS J101152.98+544206.4 (J1011+5442 for short) gradually returns to the type 1 state after a short period between 2014 and 2019 in the faint type 1.9 state. Methods. Motivated by the rebrightening in optical and mid-infrared light curves from ZTF and WISE, we obtained the new spectroscopic observations by Xinglong 2.16-m, Lijiang 2.4-m, and MMT 6.5-m optical telescopes in 2024. Results. After changing the optical AGN type from 1 to 1.9 between 2003 and 2015 based on the repeat spectroscopy from the Time Domain Spectroscopic Survey, J1011+5442 returns to its type 1 state in 2024. We detect the significant and very broad Hbeta lines (FWHM > 5000 km/s) based on the new spectra, which suggests that J1011+5442 is in the intermediate state between the dim state in 2015 and the bright state in 2003. The long-term optical and mid-infrared light curves also show a brightening trend between 2019 and 2024 as the broad Hbeta line appears. The time lag of about 100 days between the mid-infrared and optical variability is consistent with the prediction of dust reverberation mapping. Conclusions. The behaviors of the photometric and spectroscopic observations of J1011+5442 are consistent with the argument that the repeating changing-look phenomenon is regulated by the variation of accretion rate., Comment: 8 pages, 6 figures, accepted by A&A

Integrated photonic microresonators have become an essential resource for generating photonic qubits for quantum information processing, entanglement distribution and networking, and quantum communications. The pair generation rate is enhanced by reducing the microresonator radius, but this comes at the cost of increasing the frequency mode spacing and reducing the quantum information spectral density. Here, we circumvent this rate-density trade-off in an AlGaAs-on-insulator photonic device by multiplexing an array of 20 small-radius microresonators each producing a 650-GHz-spaced comb of time-energy entangled-photon pairs. The resonators can be independently tuned via integrated thermo-optic heaters, enabling control of the mode spacing from degeneracy up to a full free spectral range. We demonstrate simultaneous pumping of five resonators with up to $50$ GHz relative comb offsets, where each resonator produces pairs exhibiting time-energy entanglement visibilities up to 95$\%$, coincidence-to-accidental ratios exceeding 5,000, and an on-chip pair rate up to 2.6 GHz/mW$^2$ per comb line -- more than 40 times improvement over prior work. As a demonstration, we generate frequency-bin qubits in a maximally entangled two-qubit Bell state with fidelity exceeding 87$\%$ (90$\%$ with background correction) and detected frequency-bin entanglement rates up to 7 kHz ($\sim 70$ MHz on-chip pair rate) using $\sim 250$ $\mu$W pump power. Multiplexing small-radius microresonators combines the key capabilities required for programmable and dense photonic qubit encoding while retaining high pair-generation rates, heralded single-photon purity, and entanglement fidelity.

We demonstrate that by using 1-m-long gas-filled hollow capillary fiber (HCF) with a core diameter of 100 {\mu}m, tunable ultraviolet (UV) dispersive-wave (DW) pulses can be generated in a compact, single-stage set-up driven directly by 40-fs Ti: sapphire laser pulses. By adjusting the gas type and pressure inside the HCF, the central wavelength of the UV DW can be continuously tuned from 185 nm to ~450 nm. In the experiment, we found that for longer-wavelength (from ~320 to ~450 nm) DW generation, Raman-active gas filled in the HCF can efficiently suppress the pulse splitting effect of the high-order soliton due to the Raman-induced pulse energy dissipation, leading to the high-quality DW generation at these wavelengths with smooth, single-peak spectra. These results provide some useful insights for designing compact, wavelength-tunable ultrafast UV light sources with microjoule-level pulse energies.

Autoregressive models, built based on the Next Token Prediction (NTP) paradigm, show great potential in developing a unified framework that integrates both language and vision tasks. In this work, we rethink the NTP for autoregressive image generation and propose a novel Next Patch Prediction (NPP) paradigm. Our key idea is to group and aggregate image tokens into patch tokens containing high information density. With patch tokens as a shorter input sequence, the autoregressive model is trained to predict the next patch, thereby significantly reducing the computational cost. We further propose a multi-scale coarse-to-fine patch grouping strategy that exploits the natural hierarchical property of image data. Experiments on a diverse range of models (100M-1.4B parameters) demonstrate that the next patch prediction paradigm could reduce the training cost to around 0.6 times while improving image generation quality by up to 1.0 FID score on the ImageNet benchmark. We highlight that our method retains the original autoregressive model architecture without introducing additional trainable parameters or specifically designing a custom image tokenizer, thus ensuring flexibility and seamless adaptation to various autoregressive models for visual generation., Comment: Code: https://github.com/PKU-YuanGroup/Next-Patch-Prediction, v2: add related work "Patch-Level Training for Large Language Models"

Current efforts to learn scalable policies in robotic manipulation primarily fall into two categories: one focuses on "action," which involves behavior cloning from extensive collections of robotic data, while the other emphasizes "vision," enhancing model generalization by pre-training representations or generative models, also referred to as world models, using large-scale visual datasets. This paper presents an end-to-end paradigm that predicts actions using inverse dynamics models conditioned on the robot's forecasted visual states, named Predictive Inverse Dynamics Models (PIDM). By closing the loop between vision and action, the end-to-end PIDM can be a better scalable action learner. In practice, we use Transformers to process both visual states and actions, naming the model Seer. It is initially pre-trained on large-scale robotic datasets, such as DROID, and can be adapted to realworld scenarios with a little fine-tuning data. Thanks to large-scale, end-to-end training and the synergy between vision and action, Seer significantly outperforms previous methods across both simulation and real-world experiments. It achieves improvements of 13% on the LIBERO-LONG benchmark, 21% on CALVIN ABC-D, and 43% in real-world tasks. Notably, Seer sets a new state-of-the-art on CALVIN ABC-D benchmark, achieving an average length of 4.28, and exhibits superior generalization for novel objects, lighting conditions, and environments under high-intensity disturbances on real-world scenarios. Code and models are publicly available at https://github.com/OpenRobotLab/Seer/., Comment: Project page: https://nimolty.github.io/Seer/

We uncover a phenomenon largely overlooked by the scientific community utilizing AI: neural networks exhibit high susceptibility to minute perturbations, resulting in significant deviations in their outputs. Through an analysis of five diverse application areas -- weather forecasting, chemical energy and force calculations, fluid dynamics, quantum chromodynamics, and wireless communication -- we demonstrate that this vulnerability is a broad and general characteristic of AI systems. This revelation exposes a hidden risk in relying on neural networks for essential scientific computations, calling further studies on their reliability and security., Comment: 26 pages, 6 figures

We report the search for neutrinoless double-beta decay of $^{136}$Xe from the PandaX-4T experiment with a 3.7-tonne natural xenon target. The data reconstruction and the background modeling are optimized in the MeV energy region. A blind analysis is performed with data from the commissioning run and the first science run. No significant excess of signal over the background is observed. A lower limit on the half-life of $^{136}$Xe neutrinoless double-beta decay is established to be $2.1 \times 10^{24}$~yr at the 90\% confidence level, with a $^{136}$Xe exposure of 44.6~kg$\cdot$year. Our result represents the most stringent constraint from a natural xenon detector to date., Comment: 9 pages, 4 figures, 2 tables

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

In this paper, we introduce ManiVideo, a novel method for generating consistent and temporally coherent bimanual hand-object manipulation videos from given motion sequences of hands and objects. The core idea of ManiVideo is the construction of a multi-layer occlusion (MLO) representation that learns 3D occlusion relationships from occlusion-free normal maps and occlusion confidence maps. By embedding the MLO structure into the UNet in two forms, the model enhances the 3D consistency of dexterous hand-object manipulation. To further achieve the generalizable grasping of objects, we integrate Objaverse, a large-scale 3D object dataset, to address the scarcity of video data, thereby facilitating the learning of extensive object consistency. Additionally, we propose an innovative training strategy that effectively integrates multiple datasets, supporting downstream tasks such as human-centric hand-object manipulation video generation. Through extensive experiments, we demonstrate that our approach not only achieves video generation with plausible hand-object interaction and generalizable objects, but also outperforms existing SOTA methods.

Benefiting from the significantly improved energy density and safety, all-solid-state lithium batteries (ASSLBs) are considered one of the most promising next-generation energy technologies. Their practical applications, however, are strongly impeded by the Li dendrite formation. Despite this recognized challenge, a comprehensive understanding of Li dendrite nucleation and formation mechanism remains elusive. In particular, the initial locations of Li dendrite formation are still ambiguous: do Li clusters form directly at the Li anode surface, or inside the bulk solid electrolyte (SE), or within the solid-electrolyte interphase (SEI)? Here, based on the deep-potential molecular dynamics simulations combined with enhanced sampling techniques, we investigate the atomic-level mechanism of Li cluster nucleation and formation at the Li anode/SE interface. We observe that an isolated Li cluster initially forms inside the SEI between the Li6PS5Cl SE and the Li metal anode, located ~1 nm away from the Li anode/SEI boundary. The local electronic structure of the spontaneously formed SEI is found to be a key factor enabling the Li cluster formation within SEI, in which a significantly decreased bandgap could facilitate electronic conduction through the SEI and reduce Li+ ions to metallic Li atoms therein. Our work therefore provides atomic-level insights into Li-dendrite nucleation at anode/SE interfaces in ASSLBs, and could guide future design for developing Li-dendrite-inhibiting strategies.

With the integration of large language models (LLMs), embodied agents have strong capabilities to execute complicated instructions in natural language, paving a way for the potential deployment of embodied robots. However, a foreseeable issue is that those embodied agents can also flawlessly execute some hazardous tasks, potentially causing damages in real world. To study this issue, we present SafeAgentBench -- a new benchmark for safety-aware task planning of embodied LLM agents. SafeAgentBench includes: (1) a new dataset with 750 tasks, covering 10 potential hazards and 3 task types; (2) SafeAgentEnv, a universal embodied environment with a low-level controller, supporting multi-agent execution with 17 high-level actions for 8 state-of-the-art baselines; and (3) reliable evaluation methods from both execution and semantic perspectives. Experimental results show that the best-performing baseline gets 69% success rate for safe tasks, but only 5% rejection rate for hazardous tasks, indicating significant safety risks. More details and codes are available at https://github.com/shengyin1224/SafeAgentBench., Comment: 21 pages, 14 tables, 7 figures, submitted to ICRA 2024

Link prediction in dynamic graphs (LPDG) has been widely applied to real-world applications such as website recommendation, traffic flow prediction, organizational studies, etc. These models are usually kept local and secure, with only the interactive interface restrictively available to the public. Thus, the problem of the black-box evasion attack on the LPDG model, where model interactions and data perturbations are restricted, seems to be essential and meaningful in practice. In this paper, we propose the first practicable black-box evasion attack method that achieves effective attacks against the target LPDG model, within a limited amount of interactions and perturbations. To perform effective attacks under limited perturbations, we develop a graph sequential embedding model to find the desired state embedding of the dynamic graph sequences, under a deep reinforcement learning framework. To overcome the scarcity of interactions, we design a multi-environment training pipeline and train our agent for multiple instances, by sharing an aggregate interaction buffer. Finally, we evaluate our attack against three advanced LPDG models on three real-world graph datasets of different scales and compare its performance with related methods under the interaction and perturbation constraints. Experimental results show that our attack is both effective and practicable.

In real-world applications across specialized domains, addressing complex out-of-distribution (OOD) challenges is a common and significant concern. In this study, we concentrate on the task of fine-grained bronze ware dating, a critical aspect in the study of ancient Chinese history, and developed a benchmark dataset named ShiftedBronzes. By extensively expanding the bronze Ding dataset, ShiftedBronzes incorporates two types of bronze ware data and seven types of OOD data, which exhibit distribution shifts commonly encountered in bronze ware dating scenarios. We conduct benchmarking experiments on ShiftedBronzes and five commonly used general OOD datasets, employing a variety of widely adopted post-hoc, pre-trained Vision Large Model (VLM)-based and generation-based OOD detection methods. Through analysis of the experimental results, we validate previous conclusions regarding post-hoc, VLM-based, and generation-based methods, while also highlighting their distinct behaviors on specialized datasets. These findings underscore the unique challenges of applying general OOD detection methods to domain-specific tasks such as bronze ware dating. We hope that the ShiftedBronzes benchmark provides valuable insights into both the field of bronze ware dating and the and the development of OOD detection methods. The dataset and associated code will be available later., Comment: 9pages, 7 figures, 4 tables

Despite the rapid advancements in text-to-image (T2I) synthesis, enabling precise visual control remains a significant challenge. Existing works attempted to incorporate multi-facet controls (text and sketch), aiming to enhance the creative control over generated images. However, our pilot study reveals that the expressive power of humans far surpasses the capabilities of current methods. Users desire a more versatile approach that can accommodate their diverse creative intents, ranging from controlling individual subjects to manipulating the entire scene composition. We present VersaGen, a generative AI agent that enables versatile visual control in T2I synthesis. VersaGen admits four types of visual controls: i) single visual subject; ii) multiple visual subjects; iii) scene background; iv) any combination of the three above or merely no control at all. We train an adaptor upon a frozen T2I model to accommodate the visual information into the text-dominated diffusion process. We introduce three optimization strategies during the inference phase of VersaGen to improve generation results and enhance user experience. Comprehensive experiments on COCO and Sketchy validate the effectiveness and flexibility of VersaGen, as evidenced by both qualitative and quantitative results., Comment: The paper has been accepted by AAAI 2025. Paper code: https://github.com/FelixChan9527/VersaGen_official

Let $(X,\omega)$ be a compact Hermitian manifold and let $\{\beta\}\in H^{1,1}(X,\mathbb R)$ be a real $(1,1)$-class with a smooth representative $\beta$, such that $\int_X\beta^n>0$. Assume that there is a bounded $\beta$-plurisubharmonic function $\rho$ on $X$. First, we provide a criterion for the weak convergence of non-pluripolar complex Monge-Amp\`ere measures associated to a sequence of $\beta$-plurisubharmonic functions. Second, this criterion is utilized to solve a degenerate complex Monge-Amp\`ere equation with an $L^1$-density. Finally, an $L^\infty$-estimate of the solution to the complex Monge-Amp\`ere equation for a finite positive Radon measure is given., Comment: 36pages, Comments welcome!

In this work, we studied the $Z_{c}(3900)$ state within the framework of effective field theory. We firstly show the construction of the Lagrangian describing meson-meson-meson and meson-diquark-diquark interactions. By using the Feynman rule, we calculate the effective potentials corresponding to the coupled channels of $D\bar{D}^{*}/D^{*}\bar{D}$ and $S_{cq}\bar{A}_{cq}/A_{cq}\bar{S}_{cq}$ with $S_{cq}$ ($A_{cq}$) the scalar (axial vector) diquark composed of $c$ and $q$ quarks. After solving the Bethe-Salpeter equation of the on-shell parametrized form and compare our numerical results with the experimental mass and width of $Z_{c}(3900)$, we find that the $Z_{c}(3900)$ state can be explained as the mixture of $D\bar{D}^{*}/D^{*}\bar{D}$ and $S_{cq}\bar{A}_{cq}/A_{cq}\bar{S}_{cq}$ components., Comment: 7 pages, 1 figure

In the era of Large Language Models (LLMs), generative linguistic steganography has become a prevalent technique for hiding information within model-generated texts. However, traditional steganography methods struggle to effectively align steganographic texts with original model-generated texts due to the lower entropy of the predicted probability distribution of LLMs. This results in a decrease in embedding capacity and poses challenges for decoding stegos in real-world communication channels. To address these challenges, we propose a semantic steganography framework based on LLMs, which construct a semantic space and map secret messages onto this space using ontology-entity trees. This framework offers robustness and reliability for transmission in complex channels, as well as resistance to text rendering and word blocking. Additionally, the stegos generated by our framework are indistinguishable from the covers and achieve a higher embedding capacity compared to state-of-the-art steganography methods, while producing higher quality stegos.

Foundation models are deep learning models pre-trained on large amounts of data which are capable of generalizing to multiple datasets and/or downstream tasks. This work demonstrates how data collected by the CMS experiment at the Large Hadron Collider can be useful in pre-training foundation models for HEP. Specifically, we introduce the AspenOpenJets dataset, consisting of approximately 180M high $p_T$ jets derived from CMS 2016 Open Data. We show how pre-training the OmniJet-$\alpha$ foundation model on AspenOpenJets improves performance on generative tasks with significant domain shift: generating boosted top and QCD jets from the simulated JetClass dataset. In addition to demonstrating the power of pre-training of a jet-based foundation model on actual proton-proton collision data, we provide the ML-ready derived AspenOpenJets dataset for further public use., Comment: 11 pages, 4 figures, the AspenOpenJets dataset can be found at http://doi.org/10.25592/uhhfdm.16505

Digital agents are increasingly employed to automate tasks in interactive digital environments such as web pages, software applications, and operating systems. While text-based agents built on Large Language Models (LLMs) often require frequent updates due to platform-specific APIs, visual agents leveraging Multimodal Large Language Models (MLLMs) offer enhanced adaptability by interacting directly with Graphical User Interfaces (GUIs). However, these agents face significant challenges in visual perception, particularly when handling high-resolution, visually complex digital environments. This paper introduces Iris, a foundational visual agent that addresses these challenges through two key innovations: Information-Sensitive Cropping (ISC) and Self-Refining Dual Learning (SRDL). ISC dynamically identifies and prioritizes visually dense regions using a edge detection algorithm, enabling efficient processing by allocating more computational resources to areas with higher information density. SRDL enhances the agent's ability to handle complex tasks by leveraging a dual-learning loop, where improvements in referring (describing UI elements) reinforce grounding (locating elements) and vice versa, all without requiring additional annotated data. Empirical evaluations demonstrate that Iris achieves state-of-the-art performance across multiple benchmarks with only 850K GUI annotations, outperforming methods using 10x more training data. These improvements further translate to significant gains in both web and OS agent downstream tasks.

This paper introduces the human-like embodied AI interviewer which integrates android robots equipped with advanced conversational capabilities, including attentive listening, conversational repairs, and user fluency adaptation. Moreover, it can analyze and present results post-interview. We conducted a real-world case study at SIGDIAL 2024 with 42 participants, of whom 69% reported positive experiences. This study demonstrated the system's effectiveness in conducting interviews just like a human and marked the first employment of such a system at an international conference. The demonstration video is available at https://youtu.be/jCuw9g99KuE., Comment: This paper has been accepted for demonstration presentation at International Conference on Computational Linguistics (COLING 2025)

Personalized diffusion models, capable of synthesizing highly realistic images based on a few reference portraits, pose substantial social, ethical, and legal risks by enabling identity replication. Existing defense mechanisms rely on computationally intensive adversarial perturbations tailored to individual images, rendering them impractical for real-world deployment. This study introduces Real-time Identity Defender (RID), a neural network designed to generate adversarial perturbations through a single forward pass, bypassing the need for image-specific optimization. RID achieves unprecedented efficiency, with defense times as low as 0.12 seconds on a single GPU (4,400 times faster than leading methods) and 1.1 seconds per image on a standard Intel i9 CPU, making it suitable for edge devices such as smartphones. Despite its efficiency, RID matches state-of-the-art performance across visual and quantitative benchmarks, effectively mitigating identity replication risks. Our analysis reveals that RID's perturbations mimic the efficacy of traditional defenses while exhibiting properties distinct from natural noise, such as Gaussian perturbations. To enhance robustness, we extend RID into an ensemble framework that integrates multiple pre-trained text-to-image diffusion models, ensuring resilience against black-box attacks and post-processing techniques, including JPEG compression and diffusion-based purification., Comment: 21 pages, 7 figures

The tremendous success of behavior cloning (BC) in robotic manipulation has been largely confined to tasks where demonstrations can be effectively collected through human teleoperation. However, demonstrations for contact-rich manipulation tasks that require complex coordination of multiple contacts are difficult to collect due to the limitations of current teleoperation interfaces. We investigate how to leverage model-based planning and optimization to generate training data for contact-rich dexterous manipulation tasks. Our analysis reveals that popular sampling-based planners like rapidly exploring random tree (RRT), while efficient for motion planning, produce demonstrations with unfavorably high entropy. This motivates modifications to our data generation pipeline that prioritizes demonstration consistency while maintaining solution diversity. Combined with a diffusion-based goal-conditioned BC approach, our method enables effective policy learning and zero-shot transfer to hardware for two challenging contact-rich manipulation tasks.

The effectiveness of large language models (LLMs) is closely tied to the design of prompts, making prompt optimization essential for enhancing their performance across a wide range of tasks. Many existing approaches to automating prompt engineering rely exclusively on textual feedback, refining prompts based solely on inference errors identified by large, computationally expensive LLMs. Unfortunately, smaller models struggle to generate high-quality feedback, resulting in complete dependence on large LLM judgment. Moreover, these methods fail to leverage more direct and finer-grained information, such as gradients, due to operating purely in text space. To this end, we introduce GReaTer, a novel prompt optimization technique that directly incorporates gradient information over task-specific reasoning. By utilizing task loss gradients, GReaTer enables self-optimization of prompts for open-source, lightweight language models without the need for costly closed-source LLMs. This allows high-performance prompt optimization without dependence on massive LLMs, closing the gap between smaller models and the sophisticated reasoning often needed for prompt refinement. Extensive evaluations across diverse reasoning tasks including BBH, GSM8k, and FOLIO demonstrate that GReaTer consistently outperforms previous state-of-the-art prompt optimization methods, even those reliant on powerful LLMs. Additionally, GReaTer-optimized prompts frequently exhibit better transferability and, in some cases, boost task performance to levels comparable to or surpassing those achieved by larger language models, highlighting the effectiveness of prompt optimization guided by gradients over reasoning. Code of GReaTer is available at https://github.com/psunlpgroup/GreaTer., Comment: 32 pages, 8 figures

Seasonal time series exhibit intricate long-term dependencies, posing a significant challenge for accurate future prediction. This paper introduces the Multi-scale Seasonal Decomposition Model (MSSD) for seasonal time-series forecasting. Initially, leveraging the inherent periodicity of seasonal time series, we decompose the univariate time series into three primary components: Ascending, Peak, and Descending. This decomposition approach enhances the capture of periodic features. By addressing the limitations of existing time-series modeling methods, particularly in modeling the Peak component, this research proposes a multi-scale network structure designed to effectively capture various potential peak fluctuation patterns in the Peak component. This study integrates Conv2d and Temporal Convolutional Networks to concurrently capture global and local features. Furthermore, we incorporate multi-scale reshaping to augment the modeling capacity for peak fluctuation patterns. The proposed methodology undergoes validation using three publicly accessible seasonal datasets. Notably, in both short-term and long-term fore-casting tasks, our approach exhibits a 10$\%$ reduction in error compared to the baseline models.