arXiv subjects

Jing Yang

Publications and source records attributed to Jing Yang.

At least 145 records · Page 8Linked to original sources

arXiv2025

HiVA: Self-organized Hierarchical Variable Agent via Goal-driven Semantic-Topological Evolution

Autonomous agents play a crucial role in advancing Artificial General Intelligence, enabling problem decomposition and tool orchestration through Large Language Models (LLMs). However, existing paradigms face a critical trade-off. On one hand, reusable fixed workflows require manual reconfiguration upon environmental changes; on the other hand, flexible reactive loops fail to distill reasoning progress into transferable structures. We introduce Hierarchical Variable Agent (HiVA), a novel framework modeling agentic workflows as self-organized graphs with the Semantic-Topological Evolution (STEV) algorithm, which optimizes hybrid semantic-topological spaces using textual gradients as discrete-domain surrogates for backpropagation. The iterative process comprises Multi-Armed Bandit-infused forward routing, diagnostic gradient generation from environmental feedback, and coordinated updates that co-evolve individual semantics and topology for collective optimization in unknown environments. Experiments on dialogue, coding, Long-context Q&A, mathematical, and agentic benchmarks demonstrate improvements of 5-10% in task accuracy and enhanced resource efficiency over existing baselines, establishing HiVA's effectiveness in autonomous task execution.

cs.AI↗

arXiv2025

EMPOWER: Evolutionary Medical Prompt Optimization With Reinforcement Learning

Prompt engineering significantly influences the reliability and clinical utility of Large Language Models (LLMs) in medical applications. Current optimization approaches inadequately address domain-specific medical knowledge and safety requirements. This paper introduces EMPOWER, a novel evolutionary framework that enhances medical prompt quality through specialized representation learning, multi-dimensional evaluation, and structure-preserving algorithms. Our methodology incorporates: (1) a medical terminology attention mechanism, (2) a comprehensive assessment architecture evaluating clarity, specificity, clinical relevance, and factual accuracy, (3) a component-level evolutionary algorithm preserving clinical reasoning integrity, and (4) a semantic verification module ensuring adherence to medical knowledge. Evaluation across diagnostic, therapeutic, and educational tasks demonstrates significant improvements: 24.7% reduction in factually incorrect content, 19.6% enhancement in domain specificity, and 15.3% higher clinician preference in blinded evaluations. The framework addresses critical challenges in developing clinically appropriate prompts, facilitating more responsible integration of LLMs into healthcare settings.

cs.CL↗

arXiv2025

CURE: Critical-Token-Guided Re-Concatenation for Entropy-Collapse Prevention

Recent advances in Reinforcement Learning with Verified Reward (RLVR) have driven the emergence of more sophisticated cognitive behaviors in large language models (LLMs), thereby enhancing their reasoning capabilities. However, in prior RLVR pipelines, the repeated use of static initial-state sampling drawn exactly from the dataset distribution during each sampling phase produced overly deterministic, low diversity model behavior, which manifested as rapid entropy collapse and hindered sustained performance gains during prolonged training. To address this issue, we introduce CURE (Critical-token-gUided Re concatenation for Entropy-collapse prevention), a two-stage framework that balances exploration and exploitation. Specifically, in the first stage, to deliberately steer the model toward novel yet coherent contexts, we re-generate at high-entropy critical tokens and jointly optimize the original and the branched trajectories. The further comparison with vanilla DAPO shows that the regeneration process achieves a better performance on math reasoning tasks while sustaining a high-level entropy degree for exploration. In the second stage, we continue training with static initial-state sampling by DAPO, intentionally placing the model in a familiar state to gradually strengthen exploitation. Extensive experiments on Qwen-2.5-Math-7B show that, compared to other RLVR methods, CURE achieves a 5% performance gain across six math benchmarks, establishing state-of-the-art performance in both entropy and accuracy. A series of experiments further validate the effectiveness of our approach. Code is available at https://github.com/bytedance/CURE.

cs.LG↗

arXiv2025

Localized Gaussian Splatting Editing with Contextual Awareness

Recent text-guided generation of individual 3D object has achieved great success using diffusion priors. However, these methods are not suitable for object insertion and replacement tasks as they do not consider the background, leading to illumination mismatches within the environment. To bridge the gap, we introduce an illumination-aware 3D scene editing pipeline for 3D Gaussian Splatting (3DGS) representation. Our key observation is that inpainting by the state-of-the-art conditional 2D diffusion model is consistent with background in lighting. To leverage the prior knowledge from the well-trained diffusion models for 3D object generation, our approach employs a coarse-to-fine objection optimization pipeline with inpainted views. In the first coarse step, we achieve image-to-3D lifting given an ideal inpainted view. The process employs 3D-aware diffusion prior from a view-conditioned diffusion model, which preserves illumination present in the conditioning image. To acquire an ideal inpainted image, we introduce an Anchor View Proposal (AVP) algorithm to find a single view that best represents the scene illumination in target region. In the second Texture Enhancement step, we introduce a novel Depth-guided Inpainting Score Distillation Sampling (DI-SDS), which enhances geometry and texture details with the inpainting diffusion prior, beyond the scope of the 3D-aware diffusion prior knowledge in the first coarse step. DI-SDS not only provides fine-grained texture enhancement, but also urges optimization to respect scene lighting. Our approach efficiently achieves local editing with global illumination consistency without explicitly modeling light transport. We demonstrate robustness of our method by evaluating editing in real scenes containing explicit highlight and shadows, and compare against the state-of-the-art text-to-3D editing methods.

cs.CV↗

arXiv2025

Best Foot Forward: Robust Foot Reconstruction in-the-wild

Accurate 3D foot reconstruction is crucial for personalized orthotics, digital healthcare, and virtual fittings. However, existing methods struggle with incomplete scans and anatomical variations, particularly in self-scanning scenarios where user mobility is limited, making it difficult to capture areas like the arch and heel. We present a novel end-to-end pipeline that refines Structure-from-Motion (SfM) reconstruction. It first resolves scan alignment ambiguities using SE(3) canonicalization with a viewpoint prediction module, then completes missing geometry through an attention-based network trained on synthetically augmented point clouds. Our approach achieves state-of-the-art performance on reconstruction metrics while preserving clinically validated anatomical fidelity. By combining synthetic training data with learned geometric priors, we enable robust foot reconstruction under real-world capture conditions, unlocking new opportunities for mobile-based 3D scanning in healthcare and retail.

cs.CV↗

arXiv2025

Pretrained Reversible Generation as Unsupervised Visual Representation Learning

Recent generative models based on score matching and flow matching have significantly advanced generation tasks, but their potential in discriminative tasks remains underexplored. Previous approaches, such as generative classifiers, have not fully leveraged the capabilities of these models for discriminative tasks due to their intricate designs. We propose Pretrained Reversible Generation (PRG), which extracts unsupervised representations by reversing the generative process of a pretrained continuous generation model. PRG effectively reuses unsupervised generative models, leveraging their high capacity to serve as robust and generalizable feature extractors for downstream tasks. This framework enables the flexible selection of feature hierarchies tailored to specific downstream tasks. Our method consistently outperforms prior approaches across multiple benchmarks, achieving state-of-the-art performance among generative model based methods, including 78% top-1 accuracy on ImageNet at a resolution of 64*64. Extensive ablation studies, including out-of-distribution evaluations, further validate the effectiveness of our approach.PRG is available at https://github.com/opendilab/PRG.

cs.CV↗

arXiv2025

DAFMSVC: One-Shot Singing Voice Conversion with Dual Attention Mechanism and Flow Matching

Singing Voice Conversion (SVC) transfers a source singer's timbre to a target while keeping melody and lyrics. The key challenge in any-to-any SVC is adapting unseen speaker timbres to source audio without quality degradation. Existing methods either face timbre leakage or fail to achieve satisfactory timbre similarity and quality in the generated audio. To address these challenges, we propose DAFMSVC, where the self-supervised learning (SSL) features from the source audio are replaced with the most similar SSL features from the target audio to prevent timbre leakage. It also incorporates a dual cross-attention mechanism for the adaptive fusion of speaker embeddings, melody, and linguistic content. Additionally, we introduce a flow matching module for high quality audio generation from the fused features. Experimental results show that DAFMSVC significantly enhances timbre similarity and naturalness, outperforming state-of-the-art methods in both subjective and objective evaluations.

cs.SD↗

arXiv2025

HH-Codec: High Compression High-fidelity Discrete Neural Codec for Spoken Language Modeling

Discrete speech tokenization is a fundamental component in speech codecs. However, in large-scale speech-to-speech systems, the complexity of parallel streams from multiple quantizers and the computational cost of high-time-dimensional codecs pose significant challenges. In this paper, we introduce HH-Codec, a neural codec that achieves extreme compression at 24 tokens per second for 24 kHz audio while relying on single-quantizer inference. Our approach involves a carefully designed Vector Quantization space for Spoken Language Modeling, optimizing compression efficiency while minimizing information loss. Building on this, we propose an asymmetric encoder-decoder architecture (Audio-VQ-Mel-Audio) that leverages dual supervision and progressive training to enhance reconstruction stability and fidelity. HH-Codec achieves state-of-the-art performance in speech reconstruction with an ultra-low bandwidth of 0.3 kbps. We further evaluate its effectiveness in codebook utilization and generative model adaptation, with extensive ablations validating the necessity of each module. HH-Codec is available at https://github.com/opendilab/HH-Codec.

cs.SD↗

arXiv2025

Dynamic Inter-Class Confusion-Aware Encoder for Audio-Visual Fusion in Human Activity Recognition

Humans do not understand individual events in isolation; rather, they generalize concepts within classes and compare them to others. Existing audio-video pre-training paradigms only focus on the alignment of the overall audio-video modalities, without considering the reinforcement of distinguishing easily confused classes through cognitive induction and contrast during training. This paper proposes the Dynamic Inter-Class Confusion-Aware Encoder (DICCAE), an encoder that aligns audio-video representations at a fine-grained, category-level. DICCAE addresses category confusion by dynamically adjusting the confusion loss based on inter-class confusion degrees, thereby enhancing the model's ability to distinguish between similar activities. To further extend the application of DICCAE, we also introduce a novel training framework that incorporates both audio and video modalities, as well as their fusion. To mitigate the scarcity of audio-video data in the human activity recognition task, we propose a cluster-guided audio-video self-supervised pre-training strategy for DICCAE. DICCAE achieves near state-of-the-art performance on the VGGSound dataset, with a top-1 accuracy of 65.5%. We further evaluate its feature representation quality through extensive ablation studies, validating the necessity of each module.

cs.CV↗

arXiv2025

Decision Feedback In-Context Learning for Wireless Symbol Detection

Pre-trained Transformers, through in-context learning (ICL), have demonstrated exceptional capabilities to adapt to new tasks using example prompts without model update. Transformer-based wireless receivers, where prompts consist of the pilot data in the form of transmitted and received signal pairs, have shown high detection accuracy when pilot data are abundant. However, pilot information is often costly and limited in practice. In this work, we propose DEcision Feedback IN-ContExt Detection (DEFINED) as a new wireless receiver design, which bypasses channel estimation and directly performs symbol detection using the (sometimes extremely) limited pilot data. The key innovation in DEFINED is the proposed decision feedback mechanism in ICL, where we sequentially incorporate the detected symbols into the prompts as pseudo-labels to improve the detection for subsequent symbols. We further establish an error lower bound and provide theoretical insights into the model's generalization under channel distribution mismatch. Extensive experiments across a broad range of wireless settings demonstrate that a small Transformer trained with DEFINED achieves significant performance improvements over conventional methods, in some cases only needing a single pilot pair to achieve similar performance to the latter with more than 4 pilot pairs.

cs.IT↗

arXiv2025

Augmenting Online RL with Offline Data is All You Need: A Unified Hybrid RL Algorithm Design and Analysis

This paper investigates a hybrid learning framework for reinforcement learning (RL) in which the agent can leverage both an offline dataset and online interactions to learn the optimal policy. We present a unified algorithm and analysis and show that augmenting confidence-based online RL algorithms with the offline dataset outperforms any pure online or offline algorithm alone and achieves state-of-the-art results under two learning metrics, i.e., sub-optimality gap and online learning regret. Specifically, we show that our algorithm achieves a sub-optimality gap $\tilde{O}(\sqrt{1/(N_0/\mathtt{C}(π^*|ρ)+N_1}) )$, where $\mathtt{C}(π^*|ρ)$ is a new concentrability coefficient, $N_0$ and $N_1$ are the numbers of offline and online samples, respectively. For regret minimization, we show that it achieves a constant $\tilde{O}( \sqrt{N_1/(N_0/\mathtt{C}(π^{-}|ρ)+N_1)} )$ speed-up compared to pure online learning, where $\mathtt{C}(π^-|ρ)$ is the concentrability coefficient over all sub-optimal policies. Our results also reveal an interesting separation on the desired coverage properties of the offline dataset for sub-optimality gap minimization and regret minimization. We further validate our theoretical findings in several experiments in special RL models such as linear contextual bandits and Markov decision processes (MDPs).

cs.LG↗

arXiv2025

Chain-of-Thought Enhanced Shallow Transformers for Wireless Symbol Detection

Transformers have shown potential in solving wireless communication problems, particularly via in-context learning (ICL), where models adapt to new tasks through prompts without requiring model updates. However, prior ICL-based Transformer models rely on deep architectures with many layers to achieve satisfactory performance, resulting in substantial storage and computational costs. In this work, we propose CHain Of thOught Symbol dEtection (CHOOSE), a CoT-enhanced shallow Transformer framework for wireless symbol detection. By introducing autoregressive latent reasoning steps within the hidden space, CHOOSE significantly improves the reasoning capacity of shallow models (1-2 layers) without increasing model depth. This design enables lightweight Transformers to achieve detection performance comparable to much deeper models, making them well-suited for deployment on resource-constrained mobile devices. Experimental results demonstrate that our approach outperforms conventional shallow Transformers and achieves performance comparable to that of deep Transformers, while maintaining storage and computational efficiency. This represents a promising direction for implementing Transformer-based algorithms in wireless receivers with limited computational resources.

cs.LG↗

arXiv2025

Uncover Treasures in DCT: Advancing JPEG Quality Enhancement by Exploiting Latent Correlations

Joint Photographic Experts Group (JPEG) achieves data compression by quantizing Discrete Cosine Transform (DCT) coefficients, which inevitably introduces compression artifacts. Most existing JPEG quality enhancement methods operate in the pixel domain, suffering from the high computational costs of decoding. Consequently, direct enhancement of JPEG images in the DCT domain has gained increasing attention. However, current DCT-domain methods often exhibit limited performance. To address this challenge, we identify two critical types of correlations within the DCT coefficients of JPEG images. Building on this insight, we propose an Advanced DCT-domain JPEG Quality Enhancement (AJQE) method that fully exploits these correlations. The AJQE method enables the adaptation of numerous well-established pixel-domain models to the DCT domain, achieving superior performance with reduced computational complexity. Compared to the pixel-domain counterparts, the DCT-domain models derived by our method demonstrate a 0.35 dB improvement in PSNR and a 60.5% increase in enhancement throughput on average.

eess.IV↗

arXiv2025

Singular profile of free boundary of incompressible inviscid fluid with external force

This article is devoted to investigate the singular profile of the free boundary of two-dimensional incompressible inviscid fluid with external force near the stagnation point. More precisely, given an external force with some polynomial type decay close to the stagnation point, the singular profile of the free boundary at stagnation point possible are corner wave, flat and cusp singularity. Through excluding the cusp and flat singularity, we know the only singular profile is corner wave singularity, and the corner depends on the decay rate of the solution near the stagnation point. The analysis depends on the geometric method to a class of Bernoulli-type free boundary problem with given degenerate gradient function on free boundary. This work is motivated by the significant work [E. V$\breve{a}$rv$\breve{a}$ruc$\breve{a}$ and G. Weiss, Acta Math, 206, 363-403, (2011)] on Stokes conjecture to the incompressible inviscid fluid acted on by gravity.

math.AP↗

arXiv2025

Breaking the Multi-Enhancement Bottleneck: Domain-Consistent Quality Enhancement for Compressed Images

Quality enhancement methods have been widely integrated into visual communication pipelines to mitigate artifacts in compressed images. Ideally, these quality enhancement methods should perform robustly when applied to images that have already undergone prior enhancement during transmission. We refer to this scenario as multi-enhancement, which generalizes the well-known multi-generation scenario of image compression. Unfortunately, current quality enhancement methods suffer from severe degradation when applied in multi-enhancement. To address this challenge, we propose a novel adaptation method that transforms existing quality enhancement models into domain-consistent ones. Specifically, our method enhances a low-quality compressed image into a high-quality image within the natural domain during the first enhancement, and ensures that subsequent enhancements preserve this quality without further degradation. Extensive experiments validate the effectiveness of our method and show that various existing models can be successfully adapted to maintain both fidelity and perceptual quality in multi-enhancement scenarios.

eess.IV↗

arXiv2025

Leveraging data-driven weather models for improving numerical weather prediction skill through large-scale spectral nudging

Operational meteorological forecasting has long relied on physics-based numerical weather prediction (NWP) models. Recently, this landscape has faced disruption by the advent of data-driven artificial intelligence (AI)-based weather models, which offer tremendous computational performance and competitive forecasting accuracy. However, data-driven models for medium-range forecasting generally suffer from major limitations, including low effective resolution and a narrow range of predicted variables. This study illustrates the relative strengths and weaknesses of these competing paradigms using the physics-based GEM (Global Environmental Multiscale) and the AI-based GraphCast models. Analyses of their respective global predictions in physical and spectral space reveal that GraphCast-predicted large scales outperform GEM, particularly for longer lead times, even though fine scales predicted by GraphCast suffer from excessive smoothing. Building on this insight, a hybrid NWP-AI system is proposed, wherein temperature and horizontal wind components predicted by GEM are spectrally nudged toward GraphCast predictions at large scales, while GEM itself freely generates the fine-scale details critical for local predictability and weather extremes. This hybrid approach is capable of leveraging the strengths of GraphCast to enhance the prediction skill of the GEM model while generating a full suite of physically consistent forecast fields with a full power spectrum. Additionally, trajectories of tropical cyclones are predicted with enhanced accuracy without significant changes in intensity. Work is in progress for operationalization of this hybrid system at the Canadian Meteorological Centre.

physics.ao-ph↗

arXiv2025

On the Training Convergence of Transformers for In-Context Classification of Gaussian Mixtures

Although transformers have demonstrated impressive capabilities for in-context learning (ICL) in practice, theoretical understanding of the underlying mechanism that allows transformers to perform ICL is still in its infancy. This work aims to theoretically study the training dynamics of transformers for in-context classification tasks. We demonstrate that, for in-context classification of Gaussian mixtures under certain assumptions, a single-layer transformer trained via gradient descent converges to a globally optimal model at a linear rate. We further quantify the impact of the training and testing prompt lengths on the ICL inference error of the trained transformer. We show that when the lengths of training and testing prompts are sufficiently large, the prediction of the trained transformer approaches the ground truth distribution of the labels. Experimental results corroborate the theoretical findings.

cs.LG↗

arXiv2025

Fixing the Double Penalty in Data-Driven Weather Forecasting Through a Modified Spherical Harmonic Loss Function

Recent advancements in data-driven weather forecasting models have delivered deterministic models that outperform the leading operational forecast systems based on traditional, physics-based models. However, these data-driven models are typically trained with a mean squared error loss function, which causes smoothing of fine scales through a "double penalty" effect. We develop a simple, parameter-free modification to this loss function that avoids this problem by separating the loss attributable to decorrelation from the loss attributable to spectral amplitude errors. Fine-tuning the GraphCast model with this new loss function results in sharp deterministic weather forecasts, an increase of the model's effective resolution from 1,250km to 160km, improvements to ensemble spread, and improvements to predictions of tropical cyclone strength and surface wind extremes.

cs.LG↗