arXiv subjects

Jie Song

Publications and source records attributed to Jie Song.

At least 73 records · Page 4Linked to original sources

arXiv2026

HuggingR$^{4}$: A Progressive Reasoning Framework for Discovering Optimal Model Companions

Building effective LLM agents increasingly requires selecting appropriate AI models as tools from large open repositories (e.g., HuggingFace with > 2M models) based on natural language requests. Unlike invoking a fixed set of API tools, repository-scale model selection must handle massive, evolving candidates with incomplete metadata. Existing approaches incorporate full model descriptions into prompts, resulting in prompt bloat, excessive token costs, and limited scalability. To address these issues, we propose HuggingR$^4$, the first framework to recast model selection as an iterative reasoning process rather than one-shot retrieval. By synergistically integrating Reasoning, Retrieval, Refinement, and Reflection, HuggingR$^4$ progressively decomposes user intent, retrieves candidates through multi-round deliberation, refines selections via fine-grained analysis, and validates results through reflection. To facilitate rigorous evaluation, we introduce a large-scale benchmark comprising 14,399 diverse user requests across 37 task categories. Experiments demonstrate that HuggingR$^4$ achieves 92.03% workability and 82.46% reasonability-outperforming current state-of-the-art baselines by 26.51% and 33.25%, respectively, while reducing token consumption by $6.9 \times$.

cs.AI↗

arXiv2026

DeepPrep: An LLM-Powered Agentic System for Autonomous Data Preparation

Data preparation, which aims to transform heterogeneous and noisy raw tables into analysis-ready data, remains a major bottleneck in data science. Recent approaches leverage large language models (LLMs) to automate data preparation from natural language specifications. However, existing LLM-powered methods either make decisions without grounding in intermediate execution results, or rely on linear interaction processes that offer limited support for revising earlier decisions. To address these limitations, we propose DeepPrep, an LLM-powered agentic system for autonomous data preparation. DeepPrep constructs data preparation pipelines through iterative, execution-grounded interaction with an environment that materializes intermediate table states and returns runtime feedback. To overcome the limitations of linear interaction, DeepPrep organizes pipeline construction with tree-based agentic reasoning, enabling structured exploration and non-local revision based on execution feedback. To enable effective learning of such behaviors, we propose a progressive agentic training framework, together with data synthesis that supplies diverse and complex ADP tasks. Extensive experiments show that DeepPrep achieves data preparation accuracy comparable to strong closed-source models (e.g., GPT-5) while incurring 15x lower inference cost, while establishing state-of-the-art performance among open-source baselines and generalizing effectively across diverse datasets.

cs.DB↗

arXiv2026

From Rays to Projections: Better Inputs for Feed-Forward View Synthesis

Feed-forward view synthesis models predict a novel view in a single pass with minimal 3D inductive bias. Existing works encode cameras as Plücker ray maps, which tie predictions to the arbitrary world coordinate gauge and make them sensitive to small camera transformations, thereby undermining geometric consistency. In this paper, we ask what inputs best condition a model for robust and consistent view synthesis. We propose projective conditioning, which replaces raw camera parameters with a target-view projective cue that provides a stable 2D input. This reframes the task from a brittle geometric regression problem in ray space to a well-conditioned target-view image-to-image translation problem. Additionally, we introduce a masked autoencoding pretraining strategy tailored to this cue, enabling the use of large-scale uncalibrated data for pretraining. Our method shows improved fidelity and stronger cross-view consistency compared to ray-conditioned baselines on our view-consistency benchmark. It also achieves state-of-the-art quality on standard novel view synthesis benchmarks.

cs.CV↗

arXiv2025

Learning Multi-Modal Mobility Dynamics for Generalized Next Location Recommendation

The precise prediction of human mobility has produced significant socioeconomic impacts, such as location recommendations and evacuation suggestions. However, existing methods suffer from limited generalization capability: unimodal approaches are constrained by data sparsity and inherent biases, while multi-modal methods struggle to effectively capture mobility dynamics caused by the semantic gap between static multi-modal representation and spatial-temporal dynamics. Therefore, we leverage multi-modal spatial-temporal knowledge to characterize mobility dynamics for the location recommendation task, dubbed as \textbf{M}ulti-\textbf{M}odal \textbf{Mob}ility (\textbf{M}$^3$\textbf{ob}). First, we construct a unified spatial-temporal relational graph (STRG) for multi-modal representation, by leveraging the functional semantics and spatial-temporal knowledge captured by the large language models (LLMs)-enhanced spatial-temporal knowledge graph (STKG). Second, we design a gating mechanism to fuse spatial-temporal graph representations of different modalities, and propose an STKG-guided cross-modal alignment to inject spatial-temporal dynamic knowledge into the static image modality. Extensive experiments on six public datasets show that our proposed method not only achieves consistent improvements in normal scenarios but also exhibits significant generalization ability in abnormal scenarios.

cs.AI↗

arXiv2025

Learning Generalizable Hand-Object Tracking from Synthetic Demonstrations

We present a system for learning generalizable hand-object tracking controllers purely from synthetic data, without requiring any human demonstrations. Our approach makes two key contributions: (1) HOP, a Hand-Object Planner, which can synthesize diverse hand-object trajectories; and (2) HOT, a Hand-Object Tracker that bridges synthetic-to-physical transfer through reinforcement learning and interaction imitation learning, delivering a generalizable controller conditioned on target hand-object states. Our method extends to diverse object shapes and hand morphologies. Through extensive evaluations, we show that our approach enables dexterous hands to track challenging, long-horizon sequences including object re-arrangement and agile in-hand reorientation. These results represent a significant step toward scalable foundation controllers for manipulation that can learn entirely from synthetic data, breaking the data bottleneck that has long constrained progress in dexterous manipulation.

cs.RO↗

arXiv2025

Towards Enhanced Image Generation Via Multi-modal Chain of Thought in Unified Generative Models

Unified generative models have shown remarkable performance in text and image generation. For image synthesis tasks, they adopt straightforward text-to-image (T2I) generation. However, direct T2I generation limits the models in handling complex compositional instructions, which frequently occur in real-world scenarios. Although this issue is vital, existing works mainly focus on improving the basic image generation capability of the models. While such improvements help to some extent, they still fail to adequately resolve the problem. Inspired by Chain of Thought (CoT) solving complex problems step by step, this work aims to introduce CoT into unified generative models to address the challenges of complex image generation that direct T2I generation cannot effectively solve, thereby endowing models with enhanced image generation ability. To achieve this, we first propose Functionality-oriented eXperts (FoXperts), an expert-parallel architecture in our model FoX, which assigns experts by function. FoXperts disentangles potential conflicts in mainstream modality-oriented designs and provides a solid foundation for CoT. When introducing CoT, the first question is how to design it for complex image generation. To this end, we emulate a human-like artistic workflow--planning, acting, reflection, and correction--and propose the Multimodal Chain of Thought (MCoT) approach, as the data involves both text and image. To address the subsequent challenge of designing an effective MCoT training paradigm, we develop a multi-task joint training scheme that equips the model with all capabilities required for each MCoT step in a disentangled manner. This paradigm avoids the difficulty of collecting consistent multi-step data tuples. Extensive experiments show that FoX consistently outperforms existing unified models on various T2I benchmarks, delivering notable improvements in complex image generation.

cs.CV↗

arXiv2025

Dual-coding contrastive learning based on ConvNeXt and ViT models for morphological classification of galaxies in COSMOS-Web

In our previous works, we proposed a machine learning framework named \texttt{USmorph} for efficiently classifying galaxy morphology. In this study, we propose a self-supervised method called contrastive learning to upgrade the unsupervised machine learning (UML) part of the \texttt{USmorph} framework, aiming to improve the efficiency of feature extraction in this step. The upgraded UML method primarily consists of the following three aspects. (1) We employ a Convolutional Autoencoder to denoise galaxy images and the Adaptive Polar Coordinate Transformation to enhance the model's rotational invariance. (2) A pre-trained dual-encoder convolutional neural network based on ConvNeXt and ViT is used to encode the image data, while contrastive learning is then applied to reduce the dimension of the features. (3) We adopt a Bagging-based clustering model to cluster galaxies with similar features into distinct groups. By carefully dividing the redshift bins, we apply this model to the rest-frame optical images of galaxies in the COSMOS-Web field within the redshift range of $0.5 < z < 6.0$. Compared to the previous algorithm, the improved UML method successfully classifies 73\% galaxies. Using the GoogleNet algorithm, we classify the morphology of the remaining 27\% galaxies. To validate the reliability of our updated algorithm, we compared our classification results with other galaxy morphological parameters and found a good consistency with galaxy evolution. Benefiting from its higher efficiency, this updated algorithm is well-suited for application in future China Space Station Telescope missions.

astro-ph.GA↗

arXiv2025

Robustness Analysis of USmorph: I. Generalization Efficiency of Unsupervised Strategies and Supervised Learning in Galaxy Morphological Classification

We conduct a systematic robustness analysis of the hybrid machine learning framework \texttt{USmorph}, which integrates unsupervised and supervised learning for galaxy morphological classification. Although \texttt{USmorph} has already been applied to nearly 100,000 $I$-band galaxy images in the COSMOS field ($0.2 < z < 1.2$, $I_{\mathrm{mag}} < 25$), the stability of its core modules has not been quantitatively assessed. Our tests show that the convolutional autoencoder (CAE) achieves the best performance in preserving structural information when adopting an intermediate network depth, $5\times5$ convolutional kernels, and a 40-dimensional latent representation. The adaptive polar coordinate transform (APCT) effectively enhances rotational invariance and improves the robustness of downstream tasks. In the unsupervised stage, a bagging clustering number of $K=50$ provides the optimal trade-off between classification granularity and labeling efficiency. For supervised learning, we employ GoogLeNet, which exhibits stable performance without overfitting. We validate the reliability of the final classifications through two independent tests: (1) the t-distributed stochastic neighbor embedding (t-SNE) visualization reveals clear clustering boundaries in the low-dimensional space; and (2) the morphological classifications are consistent with theoretical expectations of galaxy evolution, with both true and false positives showing unbiased distributions in the parameter space. These results demonstrate the strong robustness of the \texttt{USmorph} algorithm, providing guidance for its future application to the China Space Station Telescope (CSST) mission.

astro-ph.GA↗

arXiv2025

DualMap: Online Open-Vocabulary Semantic Mapping for Natural Language Navigation in Dynamic Changing Scenes

We introduce DualMap, an online open-vocabulary mapping system that enables robots to understand and navigate dynamically changing environments through natural language queries. Designed for efficient semantic mapping and adaptability to changing environments, DualMap meets the essential requirements for real-world robot navigation applications. Our proposed hybrid segmentation frontend and object-level status check eliminate the costly 3D object merging required by prior methods, enabling efficient online scene mapping. The dual-map representation combines a global abstract map for high-level candidate selection with a local concrete map for precise goal-reaching, effectively managing and updating dynamic changes in the environment. Through extensive experiments in both simulation and real-world scenarios, we demonstrate state-of-the-art performance in 3D open-vocabulary segmentation, efficient scene mapping, and online language-guided navigation. Project page: https://eku127.github.io/DualMap/

cs.RO↗

arXiv2025

Spontaneous emission in Casimir-Rabi oscillations through a weak optomechanical coupling

The dynamical Casimir effect (DCE) describes the energy conversion from a mechanical motion to the electromagnetic fields. When the mechanical oscillator is in a mechanically excited state, the free evolution due to the DCE produces radiation in the vacuum, in analogy with the spontaneous emission from an excited atom. In this manuscript, we investigate such a spontaneous radiation process by employing the quantum trajectory approach. When the dissipation rate of the system is very low, there can be a reversible energy exchange between the mirror in the excited state and the vacuum field, and this reversible exchange is called vacuum Casimir-Rabi oscillations. Multiple quantum trajectory simulations of this process show that the number of trajectories responsible for the generation of radiation can reach a significant value when the mechanical dissipation rate is less than the photon dissipation rate. We also find that two-photon (two/three-phonon) bundle emission occurs in photon (phonon) emission. In comparison to pure two-photon and three-phonon free dissipation, the probability of two-photon bundle emission and two-phonon bundle emission are observed to be marginally elevated as a consequence of the presence of the DCE. This pattern may assist in developing a deeper comprehension of the physical characteristics of photon and phonon emission in the DCE.

quant-ph↗

arXiv2025

ProtoPFormer: Concentrating on Prototypical Parts in Vision Transformers for Interpretable Image Recognition

Prototypical part network (ProtoPNet) has drawn wide attention and boosted many follow-up studies due to its self-explanatory property for explainable artificial intelligence (XAI). However, when directly applying ProtoPNet on vision transformer (ViT) backbones, learned prototypes have a "distraction" problem: they have a relatively high probability of being activated by the background and pay less attention to the foreground. The powerful capability of modeling long-term dependency makes the transformer-based ProtoPNet hard to focus on prototypical parts, thus severely impairing its inherent interpretability. This paper proposes prototypical part transformer (ProtoPFormer) for appropriately and effectively applying the prototype-based method with ViTs for interpretable image recognition. The proposed method introduces global and local prototypes for capturing and highlighting the representative holistic and partial features of targets according to the architectural characteristics of ViTs. The global prototypes are adopted to provide the global view of objects to guide local prototypes to concentrate on the foreground while eliminating the influence of the background. Afterwards, local prototypes are explicitly supervised to concentrate on their respective prototypical visual parts, increasing the overall interpretability. Extensive experiments demonstrate that our proposed global and local prototypes can mutually correct each other and jointly make final decisions, which faithfully and transparently reason the decision-making processes associatively from the whole and local perspectives, respectively. Moreover, ProtoPFormer consistently achieves superior performance and visualization results over the state-of-the-art (SOTA) prototype-based baselines. Our code has been released at https://github.com/zju-vipa/ProtoPFormer.

cs.CV↗

arXiv2025

Syn-GRPO: Self-Evolving Data Synthesis for MLLM Perception Reasoning

RL (reinforcement learning) methods (e.g., GRPO) for MLLM (Multimodal LLM) perception ability has attracted wide research interest owing to its remarkable generalization ability. Nevertheless, existing reinforcement learning methods still face the problem of low data quality, where data samples cannot elicit diverse responses from MLLMs, thus restricting the exploration scope for MLLM reinforcement learning. Some methods attempt to mitigate this problem by imposing constraints on entropy, but none address it at its root. Therefore, to tackle this problem, this work proposes Syn-GRPO (Synthesis-GRPO), which employs an online data generator to synthesize high-quality training data with diverse responses in GRPO training. Specifically, Syn-GRPO consists of two components: (1) data server; (2) GRPO workflow. The data server synthesizes new samples from existing ones using an image generation model, featuring a decoupled and asynchronous scheme to achieve high generation efficiency. The GRPO workflow provides the data server with the new image descriptions, and it leverages a diversity reward to supervise the MLLM to predict image descriptions for synthesizing samples with diverse responses. Experiment results across three visual perception tasks demonstrate that Syn-GRPO improves the data quality by a large margin, achieving significant superior performance to existing MLLM perception methods, and Syn-GRPO presents promising potential for scaling long-term self-evolving RL. Our code is available at https://github.com/hqhQAQ/Syn-GRPO.

cs.CV↗

arXiv2025

Giant-Atom Quantum Batteries

Environmentally induced decoherence poses a fundamental challenge to quantum energy storage systems, causing irreversible energy dissipation and performance aging of quantum batteries (QBs). To address this issue, we propose a QB protocol utilizing the nonlocal coupling properties of giant atoms (GAs). In this architecture, both the QB and its charger are implemented as superconducting GAs with multiple nonlocal coupling points to a shared microwave waveguide. By engineering these atoms in a braided configuration, where their coupling paths are spatially interleaved, we show the emergence of decoherence-immune interaction dynamics. This unique geometry enables destructive interference between decoherence channels while preserving coherent energy transfer between the charger and the QB, thereby effectively suppressing the aging effects induced by waveguide-mediated dissipation. The charging properties of separated and nested coupled configurations are investigated. The results show that these two configurations underperform the braided configuration. Additionally, we propose a long-range chiral charging scheme that facilitates unidirectional energy transfer between the charger and the battery, with the capability to reverse the flow direction by modulating the applied magnetic flux. Our result provides guidelines for implementing a decoherence-resistant charging protocol and remote chiral QBs in circuits with GAs engineering.

quant-ph↗

arXiv2025

Capturing Head Avatar with Hand Contacts from a Monocular Video

Photorealistic 3D head avatars are vital for telepresence, gaming, and VR. However, most methods focus solely on facial regions, ignoring natural hand-face interactions, such as a hand resting on the chin or fingers gently touching the cheek, which convey cognitive states like pondering. In this work, we present a novel framework that jointly learns detailed head avatars and the non-rigid deformations induced by hand-face interactions. There are two principal challenges in this task. First, naively tracking hand and face separately fails to capture their relative poses. To overcome this, we propose to combine depth order loss with contact regularization during pose tracking, ensuring correct spatial relationships between the face and hand. Second, no publicly available priors exist for hand-induced deformations, making them non-trivial to learn from monocular videos. To address this, we learn a PCA basis specific to hand-induced facial deformations from a face-hand interaction dataset. This reduces the problem to estimating a compact set of PCA parameters rather than a full spatial deformation field. Furthermore, inspired by physics-based simulation, we incorporate a contact loss that provides additional supervision, significantly reducing interpenetration artifacts and enhancing the physical plausibility of the results. We evaluate our approach on RGB(D) videos captured by an iPhone. Additionally, to better evaluate the reconstructed geometry, we construct a synthetic dataset of avatars with various types of hand interactions. We show that our method can capture better appearance and more accurate deforming geometry of the face than SOTA surface reconstruction methods.

cs.CV↗

arXiv2025

RS3DBench: A Comprehensive Benchmark for 3D Spatial Perception in Remote Sensing

In this paper, we introduce a novel benchmark designed to propel the advancement of general-purpose, large-scale 3D vision models for remote sensing imagery. While several datasets have been proposed within the realm of remote sensing, many existing collections either lack comprehensive depth information or fail to establish precise alignment between depth data and remote sensing images. To address this deficiency, we present a visual Benchmark for 3D understanding of Remotely Sensed images, dubbed RS3DBench. This dataset encompasses 54,951 pairs of remote sensing images and pixel-level aligned depth maps, accompanied by corresponding textual descriptions, spanning a broad array of geographical contexts. It serves as a tool for training and assessing 3D visual perception models within remote sensing image spatial understanding tasks. Furthermore, we introduce a remotely sensed depth estimation model derived from stable diffusion, harnessing its multimodal fusion capabilities, thereby delivering state-of-the-art performance on our dataset. Our endeavor seeks to make a profound contribution to the evolution of 3D visual perception models and the advancement of geographic artificial intelligence within the remote sensing domain. The dataset, models and code will be accessed on the https://rs3dbench.github.io.

cs.CV↗

arXiv2025

A Survey on Explainable Reinforcement Learning: Concepts, Algorithms, Challenges

Reinforcement Learning (RL) is a popular machine learning paradigm where intelligent agents interact with the environment to fulfill a long-term goal. Driven by the resurgence of deep learning, Deep RL (DRL) has witnessed great success over a wide spectrum of complex control tasks. Despite the encouraging results achieved, the deep neural network-based backbone is widely deemed as a black box that impedes practitioners to trust and employ trained agents in realistic scenarios where high security and reliability are essential. To alleviate this issue, a large volume of literature devoted to shedding light on the inner workings of the intelligent agents has been proposed, by constructing intrinsic interpretability or post-hoc explainability. In this survey, we provide a comprehensive review of existing works on eXplainable RL (XRL) and introduce a new taxonomy where prior works are clearly categorized into model-explaining, reward-explaining, state-explaining, and task-explaining methods. We also review and highlight RL methods that conversely leverage human knowledge to promote learning efficiency and performance of agents while this kind of method is often ignored in XRL field. Some challenges and opportunities in XRL are discussed. This survey intends to provide a high-level summarization of XRL and to motivate future research on more effective XRL solutions. Corresponding open source codes are collected and categorized at https://github.com/Plankson/awesome-explainable-reinforcement-learning.

cs.LG↗

arXiv2025

RobustDexGrasp: Robust Dexterous Grasping of General Objects

The ability to robustly grasp a variety of objects is essential for dexterous robots. In this paper, we present a framework for zero-shot dynamic dexterous grasping using single-view visual inputs, designed to be resilient to various disturbances. Our approach utilizes a hand-centric object shape representation based on dynamic distance vectors between finger joints and object surfaces. This representation captures the local shape around potential contact regions rather than focusing on detailed global object geometry, thereby enhancing generalization to shape variations and uncertainties. To address perception limitations, we integrate a privileged teacher policy with a mixed curriculum learning approach, allowing the student policy to effectively distill grasping capabilities and explore for adaptation to disturbances. Trained in simulation, our method achieves success rates of 97.0% across 247,786 simulated objects and 94.6% across 512 real objects, demonstrating remarkable generalization. Quantitative and qualitative results validate the robustness of our policy against various disturbances.

cs.RO↗

arXiv2025

MoGA: 3D Generative Avatar Prior for Monocular Gaussian Avatar Reconstruction

We present MoGA, a novel method to reconstruct high-fidelity 3D Gaussian avatars from a single-view image. The main challenge lies in inferring unseen appearance and geometric details while ensuring 3D consistency and realism. Most previous methods rely on 2D diffusion models to synthesize unseen views; however, these generated views are sparse and inconsistent, resulting in unrealistic 3D artifacts and blurred appearance. To address these limitations, we leverage a generative avatar model, that can generate diverse 3D avatars by sampling deformed Gaussians from a learned prior distribution. Due to limited 3D training data, such a 3D model alone cannot capture all image details of unseen identities. Consequently, we integrate it as a prior, ensuring 3D consistency by projecting input images into its latent space and enforcing additional 3D appearance and geometric constraints. Our novel approach formulates Gaussian avatar creation as model inversion by fitting the generative avatar to synthetic views from 2D diffusion models. The generative avatar provides an initialization for model fitting, enforces 3D regularization, and helps in refining pose. Experiments show that our method surpasses state-of-the-art techniques and generalizes well to real-world scenarios. Our Gaussian avatars are also inherently animatable. For code, see https://zj-dong.github.io/MoGA/.

cs.CV↗