arXiv subjects

Xin Liu

Publications and source records attributed to Xin Liu.

At least 379 records · Page 21Linked to original sources

arXiv2025

Sample-efficient Unsupervised Policy Cloning from Ensemble Self-supervised Labeled Videos

Current advanced policy learning methodologies have demonstrated the ability to develop expert-level strategies when provided enough information. However, their requirements, including task-specific rewards, action-labeled expert trajectories, and huge environmental interactions, can be expensive or even unavailable in many scenarios. In contrast, humans can efficiently acquire skills within a few trials and errors by imitating easily accessible internet videos, in the absence of any other supervision. In this paper, we try to let machines replicate this efficient watching-and-learning process through Unsupervised Policy from Ensemble Self-supervised labeled Videos (UPESV), a novel framework to efficiently learn policies from action-free videos without rewards and any other expert supervision. UPESV trains a video labeling model to infer the expert actions in expert videos through several organically combined self-supervised tasks. Each task performs its duties, and they together enable the model to make full use of both action-free videos and reward-free interactions for robust dynamics understanding and advanced action prediction. Simultaneously, UPESV clones a policy from the labeled expert videos, in turn collecting environmental interactions for self-supervised tasks. After a sample-efficient, unsupervised, and iterative training process, UPESV obtains an advanced policy based on a robust video labeling model. Extensive experiments in sixteen challenging procedurally generated environments demonstrate that the proposed UPESV achieves state-of-the-art interaction-limited policy learning performance (outperforming five current advanced baselines on 12/16 tasks) without exposure to any other supervision except for videos.

cs.CV↗

arXiv2025

Galactic-scale Quasar Pairs from the Sloan Digital Sky Survey and Gaia DR3

We preform a systematic search for galactic-scale quasar pairs and small-scale (<3 arcsec) lenses using the SDSS DR16 quasar catalog and Gaia DR3. Candidate double quasars (both are unobscured) are identified as Gaia resolved pairs around spectroscopically confirmed SDSS quasars ($L_{\rm bol} > 10^{44.5} \ {\rm erg \ s^{-1}}$) at $0.5 < z \lesssim 4.5$. Gaia astrometric information and SDSS spectral decomposition are used to exclude foreground star superpositions, which dominate ($\gtrsim 80\%$ of) the pair sample. We identify $136$ double quasar candidates from $1120$ Gaia-resolved pairs after a magnitude and redshift cut of $G<20.5$ and $z>0.5$ (803 double quasars out of 2,497 pairs without any cuts applied) with separations of $\sim$ 0.3 - 3 arcsec, corresponding to projected physical separations of $\sim$ 3 - 30 kpc at the median redshift of the sample of $z = 1.7$. We estimate an overall double quasar (lens and physical pairs combined) fraction using this sample, corrected for pair-resolving completeness, of $5.7_{-0.3}^{+0.3} \times 10^{-4}$ (bootstrapping errors). This double quasar fraction increases toward smaller separations, consistent with earlier findings. We also find little redshift evolution of the double quasar fraction for the luminous SDSS sample, consistent with previous observations and simulation predictions. However, the observed fraction is lower than simulation predictions by $\sim$0.8 - 1.6 dex, suggesting a significant population of obscured quasar pairs are missed in our search. Future wide-area space missions targeting both unobscured and obscured quasar pairs at sub-arcsec resolutions will reveal this population of obscured quasar pairs, and extend to much lower AGN luminosities.

astro-ph.GA↗

arXiv2025

Viscous pressureless flows with free boundary in one space dimension: The constant viscosity case

We establish the global well-posedness of the free boundary problem of the viscous pressureless and almost pressureless heat conductive flows in one space dimension. In both cases, arbitrarily large but smooth initial data is considered, and the evolving fluid domains remain bounded for all time. In the viscous pressureless case, we are able to identify the terminal flow domain in terms of the initial data. In the viscous almost pressureless case, we construct the flow as a perturbation of the viscous pressureless flow, and establish the first result for the Navier-Stokes-Fourier system in the current setting.

math.AP↗

arXiv2025

Explanation Space: A New Perspective into Time Series Interpretability

Human understandable explanation of deep learning models is essential for various critical and sensitive applications. Unlike image or tabular data where the importance of each input feature (for the classifier's decision) can be directly projected into the input, time series distinguishable features (e.g. dominant frequency) are often hard to manifest in time domain for a user to easily understand. Additionally, most explanation methods require a baseline value as an indication of the absence of any feature. However, the notion of lack of feature, which is often defined as black pixels for vision tasks or zero/mean values for tabular data, is not well-defined in time series. Despite the adoption of explainable AI methods (XAI) from tabular and vision domain into time series domain, these differences limit the application of these XAI methods in practice. In this paper, we propose a simple yet effective method that allows a model originally trained on the time domain to be interpreted in other explanation spaces using existing methods. We suggest five explanation spaces, each of which can potentially alleviate these issues in certain types of time series. Our method can be easily integrated into existing platforms without any changes to trained models or XAI methods. The code will be released upon acceptance.

cs.LG↗

arXiv2025

TileLink: Generating Efficient Compute-Communication Overlapping Kernels using Tile-Centric Primitives

Large deep learning models have achieved state-of-the-art performance in a wide range of tasks. These models often necessitate distributed systems for efficient training and inference. The fundamental building blocks for distributed model execution are intra-layer parallel operators. The most effective approach to enhancing the performance of intra-layer parallel operators involves overlapping computation with communication. The overlapping can be achieved through either operator decomposition or kernel fusion. While decomposing operators is straightforward to implement, it often results in suboptimal performance. On the other hand, fusing communication kernels with compute kernels demands significant expertise and is error-prone. In this paper, we propose TileLink to enable efficient compilation and generation of overlapped compute-communication kernels. TileLink is composed of frontend and backend. In the frontend, TileLink decouples the design space of communication and computation, linking these two parts via tile-centric primitives. In the backend, TileLink translates these primitives into low-level communication instructions, integrating the communication and computation components to achieve overlapped execution. In experiments, TileLink achieves from $1.17\times$ to $20.76\times$ speedup to non-overlapping baseline and achieves performance comparable to state-of-the-art overlapping libraries on GPUs.

cs.DC↗

arXiv2025

ByteCheckpoint: A Unified Checkpointing System for Large Foundation Model Development

Checkpointing to preserve training states is crucial during the development of Large Foundation Models (LFMs), for training resumption upon various failures or changes in GPU resources and parallelism configurations. In addition, saved checkpoints are dispatched to evaluation tasks or transferred across different training stages (e.g., from pre-training to post-training). All these scenarios require resharding distributed checkpoints from one parallelism to another. In production environments, different LFMs are trained with various frameworks and storage backends, depending on model sizes and training scales. A high-performance checkpointing system is needed to enable efficient checkpoint management at scale throughout the lifecycle of LFM development. We introduce ByteCheckpoint, an industrial-grade checkpointing system for large-scale LFM training. ByteCheckpoint features: a parallelism-agnostic checkpoint representation that enables efficient load-time checkpoint resharding; a generic checkpoint saving/loading workflow to accommodate multiple training frameworks and support different storage backends; full-stack optimizations to ensure high I/O efficiency and scalability; a suite of monitoring tools to streamline large-scale performance analysis and bottleneck detection. Compared to existing open-source checkpointing systems [52, 58], ByteCheckpoint significantly reduces runtime checkpoint stalls, achieving an average reduction of 54.20x. For saving and loading times, ByteCheckpoint achieves improvements of up to 9.96x and 8.80x, respectively.

cs.AI↗

arXiv2025

AU-TTT: Vision Test-Time Training model for Facial Action Unit Detection

Facial Action Units (AUs) detection is a cornerstone of objective facial expression analysis and a critical focus in affective computing. Despite its importance, AU detection faces significant challenges, such as the high cost of AU annotation and the limited availability of datasets. These constraints often lead to overfitting in existing methods, resulting in substantial performance degradation when applied across diverse datasets. Addressing these issues is essential for improving the reliability and generalizability of AU detection methods. Moreover, many current approaches leverage Transformers for their effectiveness in long-context modeling, but they are hindered by the quadratic complexity of self-attention. Recently, Test-Time Training (TTT) layers have emerged as a promising solution for long-sequence modeling. Additionally, TTT applies self-supervised learning for iterative updates during both training and inference, offering a potential pathway to mitigate the generalization challenges inherent in AU detection tasks. In this paper, we propose a novel vision backbone tailored for AU detection, incorporating bidirectional TTT blocks, named AU-TTT. Our approach introduces TTT Linear to the AU detection task and optimizes image scanning mechanisms for enhanced performance. Additionally, we design an AU-specific Region of Interest (RoI) scanning mechanism to capture fine-grained facial features critical for AU detection. Experimental results demonstrate that our method achieves competitive performance in both within-domain and cross-domain scenarios.

cs.CV↗

arXiv2025

Towards Fully Automated Decision-Making Systems for Greenhouse Control: Challenges and Opportunities

Machine learning has been successful in building control policies to drive a complex system to desired states in various applications (e.g. games, robotics, etc.). To be specific, a number of parameters of policy can be automatically optimized from the observations of environment to be able to generate a sequence of decisions leading to the best performance. In this survey paper, we particularly explore such policy-learning techniques for another unique, practical use-case scenario--farming, in which critical decisions (e.g., water supply, heating, etc.) must be made in a timely manner to minimize risks (e.g., damage to plants) while maximizing the revenue (e.g., healthy crops) in the end. We first provide a broad overview of latest studies on it to identify not only domain-specific challenges but opportunities with potential solutions, some of which are suggested as promising directions for future research. Also, we then introduce our successful approach to being ranked second among 46 teams at the ''3rd Autonomous Greenhouse Challenge'' to use this specific example to discuss the lessons learned about important considerations for design to create autonomous farm-management systems.

cs.AI↗

arXiv2025

IHEval: Evaluating Language Models on Following the Instruction Hierarchy

The instruction hierarchy, which establishes a priority order from system messages to user messages, conversation history, and tool outputs, is essential for ensuring consistent and safe behavior in language models (LMs). Despite its importance, this topic receives limited attention, and there is a lack of comprehensive benchmarks for evaluating models' ability to follow the instruction hierarchy. We bridge this gap by introducing IHEval, a novel benchmark comprising 3,538 examples across nine tasks, covering cases where instructions in different priorities either align or conflict. Our evaluation of popular LMs highlights their struggle to recognize instruction priorities. All evaluated models experience a sharp performance decline when facing conflicting instructions, compared to their original instruction-following performance. Moreover, the most competitive open-source model only achieves 48% accuracy in resolving such conflicts. Our results underscore the need for targeted optimization in the future development of LMs.

cs.CL↗

arXiv2025

Look Before Leap: Look-Ahead Planning with Uncertainty in Reinforcement Learning

Model-based reinforcement learning (MBRL) has demonstrated superior sample efficiency compared to model-free reinforcement learning (MFRL). However, the presence of inaccurate models can introduce biases during policy learning, resulting in misleading trajectories. The challenge lies in obtaining accurate models due to limited diverse training data, particularly in regions with limited visits (uncertain regions). Existing approaches passively quantify uncertainty after sample generation, failing to actively collect uncertain samples that could enhance state coverage and improve model accuracy. Moreover, MBRL often faces difficulties in making accurate multi-step predictions, thereby impacting overall performance. To address these limitations, we propose a novel framework for uncertainty-aware policy optimization with model-based exploratory planning. In the model-based planning phase, we introduce an uncertainty-aware k-step lookahead planning approach to guide action selection at each step. This process involves a trade-off analysis between model uncertainty and value function approximation error, effectively enhancing policy performance. In the policy optimization phase, we leverage an uncertainty-driven exploratory policy to actively collect diverse training samples, resulting in improved model accuracy and overall performance of the RL agent. Our approach offers flexibility and applicability to tasks with varying state/action spaces and reward structures. We validate its effectiveness through experiments on challenging robotic manipulation tasks and Atari games, surpassing state-of-the-art methods with fewer interactions, thereby leading to significant performance improvements.

cs.LG↗

arXiv2025

A Framework for Uncertainty Estimation in Seismology Data Processing with Application to Extract Rayleigh Wave Dispersion Curves from Noise Cross-correlation Functions

Extracting meaningful information from large seismic datasets often requires estimating the uncertainty associated with the results for quantitative analysis. This uncertainty arises from both the raw data and the manually labeled annotations. We introduce an uncertainty estimation framework designed to calculate the uncertainty from manually labeled data. This framework can efficiently output the true posterior from large datasets. We apply the framework to extract Rayleigh wave phase velocity dispersion and compute the posterior distribution of the dispersion results. We utilize 62,899 noise cross-correlation function (NCF) data from 438 stations located in Yunnan Province and manually label the Rayleigh phase velocity dispersion curves. Dispersion curve extraction presents two key challenges: (1) Researchers typically derive dispersion curves from spectrograms in the periodvelocity domain, limiting the ability to directly study the relationship between NCFs and dispersion curves; (2) Assessing uncertainty in manually labeled data remains difficult. To address these challenges, the framework takes the NCFs as input and directly output both the dispersion values and the posterior of the dispersion values when processing the NCF data. This approach allows us to construct a flexible deep neural network (DNN) architecture that balances accuracy and computational efficiency.

physics.geo-ph↗

arXiv2025

Symmetry-Based Real-Space Framework for Realizing Flat Bands and Unveiling Nodal-Line Touchings

Flat band (FB) systems provide ideal playgrounds for studying correlation physics, whereas multi-orbital characteristics in real materials are distinguished from most simple FB models. Here, we propose a systematic and versatile framework for FB constructions in tight-binding (TB) models based on symmetric compact localized states (CLSs), integrating lattice and orbital degrees of freedom. We first demonstrate that any CLS can be symmetrized into a representation of the point group, which remains valid for high orbitals with finite spin-orbit coupling (SOC). Second, we determine the candidate CLS sites according to lattice symmetry, and simplify the hopping as a linear mapping between two Hilbert spaces: one of CLS sites and another of their adjacent sites. The existence of FBs depends on a non-empty kernel of the mapping. Finally, we distinguish eigenstates in the kernel to qualify as a CLS. To illustrate the versatility of our framework, we construct three representative FB models: one in two dimensions (2D) and the rest in three dimensions (3D). All of them lack special lattice structures and incorporate high orbitals. Notably, the 3D FBs can exhibit not only band touchings at points but also along lines, a feature of significant physical interest. For a comprehensive understanding, we derive a concise criterion for determining band touchings, which provides a natural explanation for the occurrence of both gapped and gapless FBs. By unifying symmetry principles in real space, our work offers a systematic approach to constructing FBs across diverse lattice systems. This framework opens new avenues for understanding and engineering FB systems, with potential implications for correlated quantum phenomena and exotic phases of matter.

cond-mat.mes-hall↗

arXiv2025

Cross-domain Random Pre-training with Prototypes for Reinforcement Learning

This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible. Unsupervised cross-domain Reinforcement Learning (RL) pre-training shows great potential for challenging continuous visual control but poses a big challenge. In this paper, we propose \textbf{C}ross-domain \textbf{R}andom \textbf{P}re-\textbf{T}raining with \textbf{pro}totypes (CRPTpro), a novel, efficient, and effective self-supervised cross-domain RL pre-training framework. CRPTpro decouples data sampling from encoder pre-training, proposing decoupled random collection to easily and quickly generate a qualified cross-domain pre-training dataset. Moreover, a novel prototypical self-supervised algorithm is proposed to pre-train an effective visual encoder that is generic across different domains. Without finetuning, the cross-domain encoder can be implemented for challenging downstream tasks defined in different domains, either seen or unseen. Compared with recent advanced methods, CRPTpro achieves better performance on downstream policy learning without extra training on exploration agents for data collection, greatly reducing the burden of pre-training. We conduct extensive experiments across eight challenging continuous visual-control domains, including balance control, robot locomotion, and manipulation. CRPTpro significantly outperforms the next best Proto-RL(C) on 11/12 cross-domain downstream tasks with only 54.5\% wall-clock pre-training time, exhibiting state-of-the-art pre-training performance with greatly improved pre-training efficiency.

cs.LG↗

arXiv2025

Adventurer: Exploration with BiGAN for Deep Reinforcement Learning

Recent developments in deep reinforcement learning have been very successful in learning complex, previously intractable problems. Sample efficiency and local optimality, however, remain significant challenges. To address these challenges, novelty-driven exploration strategies have emerged and shown promising potential. Unfortunately, no single algorithm outperforms all others in all tasks and most of them struggle with tasks with high-dimensional and complex observations. In this work, we propose Adventurer, a novelty-driven exploration algorithm that is based on Bidirectional Generative Adversarial Networks (BiGAN), where BiGAN is trained to estimate state novelty. Intuitively, a generator that has been trained on the distribution of visited states should only be able to generate a state coming from the distribution of visited states. As a result, novel states using the generator to reconstruct input states from certain latent representations would lead to larger reconstruction errors. We show that BiGAN performs well in estimating state novelty for complex observations. This novelty estimation method can be combined with intrinsic-reward-based exploration. Our empirical results show that Adventurer produces competitive results on a range of popular benchmark tasks, including continuous robotic manipulation tasks (e.g. Mujoco robotics) and high-dimensional image-based tasks (e.g. Atari games).

cs.LG↗

arXiv2025

Passive Heart Rate Monitoring During Smartphone Use in Everyday Life

Resting heart rate (RHR) is an important biomarker of cardiovascular health and mortality, but tracking it longitudinally generally requires a wearable device, limiting its availability. We present PHRM, a deep learning system for passive heart rate (HR) and RHR measurements during everyday smartphone use, using facial video-based photoplethysmography. Our system was developed using 225,773 videos from 495 participants and validated on 185,970 videos from 205 participants in laboratory and free-living conditions, representing the largest validation study of its kind. Compared to reference electrocardiogram, PHRM achieved a mean absolute percentage error (MAPE) < 10% for HR measurements across three skin tone groups of light, medium and dark pigmentation; MAPE for each skin tone group was non-inferior versus the others. Daily RHR measured by PHRM had a mean absolute error < 5 bpm compared to a wearable HR tracker, and was associated with known risk factors. These results highlight the potential of smartphones to enable passive and equitable heart health monitoring.

q-bio.TO↗

arXiv2025

Studying the $B^{0} \to J/ψh_{1}$ decays with $h_{1}(1170)-h_{1}(1415)$ mixing in the perturbative QCD approach

In this paper, we study the $B^{0} \to J/ψh_{1}$ decays for the first time by using perturbative QCD approach up to the presently known next-to-leading order accuracy. The vertex corrections present significant contribution to the amplitude. In the calculation, the mixing between two light axial-vector mesons $h_{1}(1170)$ and $h_{1}(1415)$ are also studied in detail. The observables including the branching ratios, polarization fractions and $CP$ asymmetries are predicted and discussed explicitly. It is found that the $B^{0} \to J/ψh_{1}$ decays have relatively large branching fractions, which are generally at the order of ${\cal O}(10^{-6}\sim10^{-3})$, and thus are possible to be observed by the LHCb and Belle-II experiments in the near future. Moreover, they are very sensitive to the mixing angle $θ$ and can be used to test the values of $θ$. In addition, some ratios between the branching fractions of $B^{0} \to J/ψh_{1}$ decays can provide much stronger constraints on $θ$ due to their relatively small theoretical errors. The $B^{0} \to J/ψh_{1}$ decays are generally dominated by the longitudinal polarization contributions, specifically, $f_{L}(B^{0} \to J/ψh_{1})>80\%$, except for the case that $θ\sim 35^\circ$ and $-55^\circ$. Unfortunately, the direct $CP$ asymmetries of $B^{0} \to J/ψh_{1}$ decays are too small to be observed soon even if the effect of $θ$ is considered. The future precise measurements on $B^{0} \to J/ψh_{1}$ decays are expected for testing these theoretical findings and exploring the interesting nature of $h_{1}(1170)$ and $h_{1}(1415)$.

hep-ph↗

arXiv2025

An experimental study of using artificial reefs as scour protection around an offshore wind monopile

Artificial reefs (ARs) are man-made structures deployed on the seabed to support benthic marine ecosystems. Their presence significantly damps the local flow and therefore can be used for scour protection of offshore wind monopiles. Although the concept appears feasible, the underlying flow-sediment process is very complex and has yet been systematically investigated. To fill in this gap, a set of fixed-bed flume tests were conducted to reveal the hydrodynamic details of two typical AR shapes (cubic and hemisphere) tightly placed around a monopile in a 3x3 pattern. In parallel, a set of live-bed tests were conducted to demonstrate the AR's efficiency in scour protection. The cubic ARs almost eliminate the downward flow on the upstream side of the monopile and reduces the wake flow by 50-80%. The hemisphere ARs also significantly weaken the wake flow but guide descending flow in front of the monopile. While cubic ARs decrease upstream and downstream scour depth by up to 100%, their edge scour can lead to ARs displacement and hence reduce scour protection. The hemisphere ARs provided less scour reduction, but also less edge scour, making them more adaptive to morphology changes. Based on these findings, an optimized AR layout was proposed.

physics.flu-dyn↗

arXiv2025

Flexible manipulation of chiral spin state by chemical bond in Mn triangular lattice magnet

This study investigates the influence of chemical bonds on the magnetic structure of materials, a less explored area compared to their effect on crystal stability. By analyzing the strength and directionality of chemical bonds using the electron localization function (ELF) and charge density difference (CDD) methods, we examine their impact on magnetic exchange interactions and magnetocrystalline anisotropy under specific interstitial conditions in Mn4X compounds. Our findings indicate that these properties can effectively modulate the magnetic ground state. This work not only elucidates the varied magnetism observed in Mn triangular lattice magnets but also proposes an approach for engineering chiral spin states through chemical bonding manipulation.

cond-mat.mtrl-sci↗