Search arXiv⌕ Search

arXiv subjects

Yang Gao

Publications and source records attributed to Yang Gao.

At least 55 records · Page 3Linked to original sources
arXiv2026

Python-Fortran Hybrid Programming to Fuse AI and Physical Models: Examples of AI-LDA in climate and weather models (Hf2pMDA_v1.0)

AI provides an unprecedented opportunity for advancing physics numerical modeling including data assimilation, which is a highly efficient and critically-important tool for advancing our understanding on Earth system and its applications. At the same time, deep incorporation of AI and physical modeling can make great driving to advance AI by injecting it rich physics from long time physics-based modeling development. However, since such physics models are conventionally coded in Fortran and AI algorithms usually are conveniently designed in Python, difficulties exist to directly incorporate AI algorithms into physics models, vice versa. Here, based on the F2PY protocol, we have developed a procedure that implements an infrastructure which conveniently conducts Hf2pMDA to form a program entity so that AI algorithms and physical models can invoke mutually. As examples, within Hf2pMDA, a climate coupled data assimilation (CDA) system is naturally upgraded to a strongly CDA (SCDA) system, and a 1 km high-resolution weather DA system is conveniently implemented within a multi-layer downscaling model that has multiscale DA in different nesting layers. In the climate SCDA system, a coupled general circulation model (CGCM) and a multiscale filtering algorithm is integrated by a Python main controller (PMC) that calls Fortran CGCM components and Weakly-CDA modules as well as a data-trained SCDA algorithm by latent space autoencoder in Python. In the high-resolution weather DA system, the downscaled model consisting of traditional Fortran DA modules in all mother domains and Python AE DA algorithm in the central child domain is integrated by a PMC that organizes these components. With convenient realization of deep incorporation of any AI algorithm and physics model, the Hf2pMDA has a great potential to make progress on both AI and scientific modeling.

physics.ao-ph↗
arXiv2026

Type I Solar Radio Bursts Modulated by Solar Flares

Type I solar radio bursts (noise storms) are persistent meter-wave nonthermal emissions above active regions, with their occurrence and proper?ties closely related to the local magnetic configuration and nonthermal electron acceleration. This study examines a type I noise storm on 24 December 2023 and its relation to flare activity. The noise-storm source was co-spatial with active region AR 3529 and showed frequency-dependent spatial dispersion. The associated M2.9 flare strongly modulated the emission, with the storm intensity decreasing at flare onset, recovering afterward, and shifting to higher frequencies. Based on multiwavelength observations, we suggest that pre-flare small-scale reconnection supplied nonthermal electrons to overlying closed magnetic struc?tures and maintained the storm. During the flare, magnetic reconnection above the active region produced bidirectional plasma ejections and type III bursts with bidirectional frequency drifts; the gradually decreasing starting frequency of these bursts may indicate an upward-moving reconnection site. The resulting magnetic reconfiguration disrupted electron trapping and suppressed the storm, whereas post-flare magnetic recovery allowed the emission to resume. These results show that flares can modulate type I noise storms through magnetic restructuring and provide insight into the generation mechanism of noise storms.

astro-ph.SR↗
arXiv2026

When Teacher Guidance Misleads: Reward-Aligned On-Policy Distillation

On-policy distillation (OPD) has recently emerged as a popular post-training paradigm for large language models (LLMs), providing an efficient way to transfer the knowledge and capabilities of teacher models into student models. However, teacher guidance on student-generated prefixes is not always reliable. Training should optimize the model to generate responses that are more likely to be correct, or equivalently, to get higher outcome rewards. But during OPD, the teacher model may provide guidance that discourages the student from moving toward correct trajectories or moves the student toward incorrect ones, which is misaligned with outcome reward. Such misaligned guidance is unreliable, as it would mislead the optimization process and ultimately degrade model performance. To mitigate misaligned teacher guidance, we propose Reward-Aligned On-Policy Distillation (RA-OPD). The key insight is to keep only trajectories whose induced updates move the student toward correct trajectories or discourage the student from moving toward incorrect ones. Specifically, for each sampled trajectory, RA-OPD checks whether its trajectory-level distillation return is consistent with its outcome reward and then filters out the misaligned trajectories. RA-OPD selects more reliable trajectories to improve student model performance without requiring additional computational cost. We evaluate RA-OPD on math and code benchmarks using models from the Qwen3 family and the DeepSeek-R1 family. Across seven math benchmarks and three code benchmarks, RA-OPD significantly outperforms standard OPD and other tested OPD variants.

cs.AI↗
arXiv2026

Learning from Hard Prompts: Difficulty-aware Advantage Amplification in Dynamic Sampling

Decoupled Clip and Dynamic Sampling Policy Optimization (DAPO) is a prominent variant of Group Relative Policy Optimization (GRPO). DAPO introduces several improvements over GRPO. Among these, Dynamic Sampling contributes the most to DAPO's accuracy gains relative to GRPO. To improve accuracy, Dynamic Sampling enhances training stability by eliminating zero policy gradients from zero advantages. Specifically, it avoids such zero gradients by filtering out prompts where sampled responses are either entirely correct or incorrect. However, our theoretical analysis shows that Dynamic Sampling decrease training efficiency as it cannot effectively utilize hard-to-sample correct responses on hard prompts. Formally, it asymmetrically amplifies the advantages of distinct responses to the same prompts. On hard prompts, incorrect responses undergo greater amplification than correct ones. This leads the model to avoid generating the observed incorrect responses rather than capitalizing on the hard-to-sample correct ones on hard prompts, resulting in low training efficiency. To improve training efficiency, we propose Direct Advantage Amplification (DAA), which amplifies the advantages of hard-to-sample correct responses on hard prompts, as obtained by Dynamic Sampling. This ensures that, when Dynamic Sampling is used, these hard-to-sample responses can be effectively capitalized on, implying higher training efficiency. By integrating DAA into DAPO, we obtain Difficulty-aware Advantage Amplification Policy Optimization (DA3PO), which is implemented with fewer than 30 lines of code from DAPO. Experiments show that DA3PO significantly outperforms GRPO and other classical GRPO variants.

cs.AI↗
arXiv2026

Coronary Mask Guided Registration for Continuous Time 4D Cardiac CT Dataset Construction

Objective: Clinical cardiac CT multiphase reconstructions generally provide acceptable image quality in end-diastole (ED) or end-systole (ES) phases, but in other phases may exhibit motion artifacts, especially in the right coronary artery (RCA). This limits ground-truth availability in 4D cardiac CT imaging research. We aim to construct a 4D cardiac CT dataset that is generally suitable to serve as pseudo ground truth. Methods: We propose Coronary Mask Guided Registration (CMGR) to produce a motion-preserved, artifact-reduced, and continuous-time 4D cardiac CT sequence from the clinical multiphase reconstruction of each patient. For artifact reduction, CMGR uses the ED or ES phase as the reference phase and warps the reference volume with deformation fields to produce the sequence. For motion preservation, CMGR registers the reference phase to each non-reference phase of the multiphase reconstruction. To capture the motion of both the RCA and other cardiac structures in each registration, CMGR regularizes RCA masks and incorporates them into image-domain registration. Time-continuity is achieved by interpolating the deformation fields for non-reference phases to arbitrary times. Results: CMGR outperformed representative image-domain registration methods in capturing RCA motion and providing reasonable RCA shape, and showed competitive performance in capturing whole-heart motion. Additionally, CMGR reduced motion artifacts from clinical multiphase reconstructions, and intermediate CMGR frames generally provided plausible transitions between discrete cardiac phases. Conclusion: CMGR provides an effective approach for constructing continuous-time 4D cardiac CT datasets. Significance: The dataset can be used in system design simulations and in reconstruction algorithm development, thereby facilitating advances in cardiac CT imaging.

eess.IV↗
arXiv2026

Understanding and Enforcing Weight Disentanglement in Task Arithmetic

Task arithmetic provides an efficient, training-free way to edit pre-trained models, yet lacks a fundamental theoretical explanation for its success. The existing concept of ``weight disentanglement" describes the ideal outcome of non-interfering task composition but does not reveal its underlying cause. Crucially, what intrinsic properties of the pre-trained model ($θ_0$) or the task vectors ($τ_t$) enable this disentanglement remains underexplored. In this paper, we introduce Task-Feature Specialization (TFS), a model's ability to allocate distinct internal features to different tasks, as the fundamental principle. We first prove that TFS is a sufficient condition for weight disentanglement. More importantly, we find that TFS also gives rise to an observable geometric consequence: weight vector orthogonality. This positions TFS as the common cause for both the desired functional outcome (disentanglement) and a measurable geometric property (orthogonality). This relationship provides the key insight for our method: since the abstract TFS property is intractable to enforce directly, we can instead promote weight disentanglement by shaping its concrete geometric consequence, orthogonality. Therefore, we propose OrthoReg, a simple and effective regularization method that actively enforces an internal orthogonal structure on weight updates ($ΔW$) that constitute $τ_t$ during fine-tuning. And we theoretically prove that OrthoReg promotes disentanglement. Extensive experiments demonstrate that OrthoReg consistently and significantly enhances the performance of various task arithmetic methods. Code is available at \href{https://github.com/RL-MIND/OrthoReg}{https://github.com/RL-MIND/OrthoReg}.

cs.AI↗
arXiv2026

Windmill Spin Dynamics and Its Induced Anomalous Hall Effect in Noncollinear Antiferromagnet Mn3Sn

We demonstrate that the spin dynamics of the noncollinear antiferromagnet Mn3Sn hosts a soft eigenmode that transitions from small-angle oscillation to a large-angle, chiral windmill precession once the canting angle exceeds a threshold set by the bending of the spin order. This windmill precession has a fixed chirality of motion, which couples to conduction electrons via a Berry connection polarizability in the mixed space of momentum and spin order, generating a nonzero Berry curvature. This Berry curvature produces a time-independent, DC anomalous Hall effect that persists throughout windmill precession, in sharp contrast to the vanishing Hall response of the static equilibrium spin order when the Hall plane coincides with the Kagome plane. Our results establish a direct link between spin group symmetry, nonlinear spin dynamics, and quantum geometric transport in noncollinear antiferromagnets.

cond-mat.mtrl-sci↗
arXiv2026

MOSS-VL Technical Report

We present MOSS-VL, an open vision-language model family that treats real-time interaction -- perceiving while it speaks -- as a first-class capability. It is co-designed across the stack: the language decoder attends to vision only through gated cross-attention, so the model can naturally see incoming frames while generating; a synthesized interaction corpus supervises when to speak, when to stay silent, and when to revise; and a staged curriculum concentrates all real-time-specific training in one light final stage over a strong offline foundation. Offline, MOSS-VL-Instruct is competitive at comparable scale and leads temporal-reasoning video sets. Across four streaming benchmarks, MOSS-VL-Realtime posts the best average on three (second on the fourth) among open-source streaming models, sweeping the three subsets that squarely test proactive behavior -- 66.0 vs. 37.5 for the best baseline on OmniMMI Proactive Alerting. With 11.3B parameters but visual tokens outside the decoded sequence, MOSS-VL widens its time-to-first-token advantage over same-backbone Qwen3-VL-8B from 2.8x to 5.1x as visual context grows. We release all five checkpoints, the training curriculum, and the real-time inference code at https://github.com/OpenMOSS/MOSS-VL.

cs.CV↗
arXiv2026

Learning from Unreachable Rewards: Hint-Conditioned Reinforcement Learning for Generative Recommendation

Semantic-ID generative recommenders represent each item as a short sequence of discrete semantic tokens and predict the next item by autoregressively generating this token sequence. This paradigm enables a unified generation interface for item IDs, histories, and item text, but it also creates a structured optimization bottleneck during reward-based post-training: when an early semantic token enters the wrong branch of the item-token space, finite rollout groups rarely reach the ground-truth item, so group-relative optimization receives identical zero rewards and produces no useful advantage. We propose Hint-Conditioned Generative Recommendation (HCGRec), a semantic-ID generative recommendation framework that recovers learning signal for such hard training instances. HCGRec diagnoses each instance with checkpoint rollouts and supplies a minimal target-prefix hint only when the current generator cannot reach the correct item. The model then generates the unhinted suffix under the hinted semantic branch, turning zero-reward groups into informative comparisons over item-token completions. Hinting also changes token identity: hinted prefix tokens are oracle-provided item context, while unhinted suffix tokens are sampled generation actions. We therefore introduce hint-aware credit decomposition, using supervised learning to preserve item-semantic and prefix-structure alignment for hinted tokens and GRPO to optimize the sampled suffix. Experiments on sequential recommendation benchmarks show that HCGRec substantially improves over supervised fine-tuning and vanilla reward-based post-training, while reducing zero-advantage training samples from over 70% to below 20%. The code is accessible at https://github.com/WncFht/GRec.

cs.IR↗
arXiv2026

VoxStruct3D: Structure-Leading Flow Matching for Voxel-Space 3D MRI Synthesis

High-fidelity 3D MRI synthesis requires both globally coherent anatomy and fine-grained voxel-level detail. Although latent diffusion makes volumetric generation tractable, its image autoencoder introduces a reconstruction bottleneck that can limit the fine detail recoverable in the final volume. We present VoxStruct3D, a voxel-space flow-matching framework that directly models full-resolution MRI volumes using a clean-data prediction objective. Its Volumetric Voxel Generator (VVG) combines factorized 3D patch embedding with overlapping upsampling, time-modulated residual refinement, and skip fusion, enabling neighboring tokens to jointly reconstruct shared voxel regions and suppress patch-boundary artifacts. To complement direct voxel-space modeling with an explicit anatomical prior, we further introduce a Structure-First, Image-Follows (SFIF) strategy. A frozen pretrained 3D medical encoder and a StructVAE extract compact structure tokens that preserve dominant anatomy, while a structure-leading schedule keeps their trajectory ahead of the image trajectory. Patch-Aligned RoPE spatially aligns the unequal token grids, and asymmetric attention enforces one-way guidance from structure to image. Experiments on pathological and healthy T1-weighted brain MRI datasets show that VoxStruct3D achieves the strongest overall performance across feature-distribution alignment, sample diversity, and perceptual quality, producing anatomically coherent and visually realistic volumes.

cs.CV↗
arXiv2026

TRCoRSurg: Temporal-Relational Co-Reasoning for Surgical Video Triplet Recognition

Understanding complex surgical scenes requires recognizing multiple interdependent entities, such as instruments, actions, and targets, while maintaining their relational consistency across time. Existing surgical triplet recognition methods struggle to jointly model intra-frame label dependencies and inter-frame temporal semantics in a unified manner. To address these limitations, we propose a unified framework that integrates spatial, relational, and temporal cues for robust surgical triplet recognition. Specifically, class-specific spatial priors are first extracted through a multi-scale encoder. These priors are then refined by a Label Correlation Modeling module with multi-scale class activation map-guided relational extraction (MS-CAMRE), enabling the model to capture both static co-occurrence patterns and dynamic contextual dependencies among triplet components. Furthermore, a Bidirectional Temporal-Relational Fusion Attention (BTRFA) module harmonizes temporal and relational representations to achieve coherent temporal reasoning. We also introduce a new evaluation metric, the Triplet Consistency Error Rate (TCER), which quantitatively measures the model's ability to preserve causal and semantic consistency across triplets. Extensive experiments on the CholecT45 and ProstaTD datasets show that our method achieves state-of-the-art performance, improving AP_IVT by 5.1 percent and 7.8 percent, respectively. Moreover, according to TCER, our approach achieves relative reductions of more than 36 percent and 25 percent on the two datasets, respectively, demonstrating the effectiveness of our framework in temporal-relational co-reasoning.

cs.CV↗
arXiv2026

MIAR: Medical Image Super-Resolution With Autoregressive Modeling

Medical Image Super-Resolution (MISR) aims to enhance spatial resolution without requiring hardware modifications. Although deep learning has yielded promising results, existing paradigms face a critical trade-off: diffusion-based methods suffer from prohibitive inference latency and compromised structural fidelity, whereas regression-based models typically produce over-smoothed results that lack perceptual realism. To address these limitations, we propose MIAR, which reformulates super-resolution as a conditional and progressive next-scale prediction task through a multi-scale autoregressive framework. To ensure structural fidelity, we augment the autoregressive backbone with a Scale-Adaptive Structural Decoder. Furthermore, we integrate a hierarchical beam search strategy during inference to mitigate the recursive error accumulation inherent in autoregressive generation, a phenomenon that is especially pronounced in medical images. Extensive experiments demonstrate that MIAR establishes new state-of-the-art benchmarks while maintaining superior fidelity. Notably, our framework achieves a 7.86% improvement in the perceptual metric MUSIQ compared with the state of the art, while simultaneously delivering a 2.02x speedup over diffusion-based methods.

cs.CV↗
arXiv2026

PCSD: Persistent Consistency for Self-Distillation in Agentic Reinforcement Learning

Large language model agents have shown strong potential in complex interactive tasks, yet their reinforcement learning (RL) is often hindered by sparse rewards, as a long multi-turn trajectory may receive only a single outcome-level signal. On-policy self-distillation (OPSD) provides dense token-level supervision from a privileged teacher, but the teacher may not be reliable at every position. Existing methods commonly rely on isolated token-level discrepancies, which can be sensitive to noise, or assign a shared step-level weight that may overlook positional variation. We propose Persistent Consistency Self-Distillation (PCSD), which derives token-level distillation weights from the local persistence of teacher-favoring signals. PCSD combines adaptive windows with exponentially decayed aggregation to capture persistent relative teacher support, applies trend-aware modulation to attenuate locally declining support, and produces continuous weights through sigmoid gating. The resulting objective is jointly optimized with GRPO, combining dense teacher guidance with sparse environmental feedback. Without inference-time skills, PCSD achieves the best ALFWorld Overall results among all baselines on both backbones, exceeding GRPO by 15.6 and 13.3 points and SDAR by 6.2 and 5.5 points, while remaining competitive on WebShop and gaining 15.8 points over GRPO on unseen ALFWorld split.

cs.AI↗
arXiv2026

GoldenRetriever: Non-Interactive Homomorphic Encrypted Retrieval for Privacy-Preserving RAG

Retrieval-Augmented Generation (RAG) enhances large language models by incorporating external knowledge, but existing pipelines typically operate on plaintext data, raising significant privacy concerns. Prior work on privacy-preserving retrieval leverages cryptographic techniques such as homomorphic encryption (HE) and private information retrieval (PIR), but often relies on interactive protocols or ranking-based selection mechanisms that incur high latency and potential information leakage. In this paper, we propose a practical non-interactive encrypted retrieval framework for RAG based on threshold selection. Instead of performing expensive top-$k$ ranking under encryption, our approach selects documents whose similarity scores exceed a predefined threshold, reducing computational complexity from quadratic to linear in the corpus size. We implement this design using CKKS-based homomorphic computation, enabling fully encrypted similarity evaluation and document selection without revealing query content, intermediate scores, or selected indices. To bridge the gap between approximate encrypted computation and discrete token reconstruction, we introduce a precision-stable mask polarization method that ensures accurate recovery of selected documents. Experiments on standard retrieval benchmarks demonstrate that our approach achieves competitive retrieval effectiveness while significantly reducing latency compared to ranking-based encrypted methods. These results highlight threshold-based selection as a practical foundation for scalable and secure RAG systems.

cs.CR↗
arXiv2026

When Does Legacy Data Start to Help? Emergent Transfer in Cross-Configuration Robot Learning

Robotic hardware evolves over time, but demonstration data is often tied to a specific sensor and actuator configuration. This raises a practical and underexplored question: when does legacy data begin to benefit an upgraded robot? We study this question on a wheeled humanoid platform across two hardware generations, where both the camera and gripper are changed while the overall morphology remains fixed. Contrary to the common assumption that more cross-configuration data is always helpful, we observe a grokking-like transition: legacy data remains ineffective until the upgraded configuration acquires a minimum level of task competence, after which co-training gains rise sharply before diminishing near saturation. We hypothesize that this task-dependent transition is governed by a transfer threshold and characterize the resulting three-phase pattern. Across real-robot manipulation tasks, we observe all three phases: no measurable benefit at low competence ($10.0\% \rightarrow 10.0\%$), a sharp gain after crossing the threshold ($23.3\% \rightarrow 86.7\%$ on flower insertion), and diminishing returns at high competence ($85.0\% \rightarrow 93.3\%$ on pen insertion). We provide a theoretical account based on gradient alignment and residual policy uncertainty, and derive a phase-aware rule for deciding when to collect more new-hardware data and when to reuse legacy demonstrations. We further validate this three-phase pattern on a mobile dual-arm watering task, with results consistent with our predictions.

cs.RO↗
arXiv2026

AgentOmnia: Scaling Agentic Models for Full-Scenario Applications

Large language model agents have advanced rapidly, yet progress remains fragmented across domains, capabilities, task difficulty, and interaction settings. We frame this as full-scenario agentic scaling and present AgentOmnia, a framework coordinating task-space definition, data synthesis, post-training, evaluation, and improvement across To-Consumer (ToC), To-Business (ToB), and To-Employee (ToE) applications. An extensible Domain x Capability x Atomic Difficulty taxonomy aligns these stages and enables fine-grained diagnosis with OmniaBench. AgentOmnia combines bidirectional environment-task synthesis with tool-dependency, program-structured, and solver-based pipelines, constructing 5,018 stateful environments with 255,375 tools and 52,361 tasks. Programs, solvers, and verifiers provide correctness signals, while supervised fine-tuning, online agentic reinforcement learning, and a rollback curriculum support post-training. Evaluation failures translate into Product Requirement Documents (PRDs) for targeted self-evolution. Starting from Qwen3-30B-A3B-Thinking-2507, AgentOmnia raises the pass rate on the OmniaBench challenging subset from 9.16% to 37.11% and the macro-average across OmniaBench, $τ^2$-Bench, DeepPlanning, and VitaBench from 22.86% to 41.69%. Under a unified protocol,it leads the evaluated agentic post-trained baselines on OmniaBench and retains the highest four-benchmark macro-average. It also surpasses Qwen3-235B-A22B-Thinking-2507 on all four benchmarks and exceeds Qwen3.5-35B-A3B on the macro-average. Gains span three application splits, ten capability dimensions, eight atomic-difficulty factors, and 76 of 90 level-1 domains, indicating broad rather than category-specific improvement. A one-round study provides initial evidence for PRD-guided self-evolution, motivating validation at larger scales and in industrial settings.

cs.AI↗
arXiv2026

Beyond Text-to-SQL: Can LLMs Really Debug Enterprise ETL SQL?

SQL is central to enterprise data engineering, yet generating fully correct SQL code in a single attempt remains difficult, even for experienced developers and advanced text-to-SQL LLMs, often requiring multiple debugging iterations. We introduce OurBench, the first benchmark for enterprise-level SQL reasoning and debugging. Our benchmark is built on two key innovations: (1) an automated construction workflow that uses reverse engineering to systematically inject realistic bugs into large-scale SQL code, enabling scalable and diverse benchmark generation; and (2) an execution-free evaluation framework tailored to enterprise settings, providing fast, accurate, and resource-efficient assessment. OurBench comprises 469 OurBenchSyn queries featuring syntax errors with explicit error messages, and 516 OurBenchSem queries targeting semantic errors in which the code fails to meet user intent. The queries are highly complex, averaging over 140 lines and featuring deep and wide abstract syntax trees. Evaluation of nearly 30 LLMs reveals a substantial performance gap: the best-performing model, Claude-4-Sonnet, achieves only 36.46 percent accuracy on OurBenchSyn and 32.17 percent on OurBenchSem, while most models score below 20 percent. We further explore four solution strategies, identify key challenges, and outline promising directions for enterprise SQL debugging with LLMs.

cs.AI↗
arXiv2026

Cross-Layer Misalignment Detection in Agent Skills: A Progressive Loading-Aware Contrastive Learning Approach

Large language model (LLM) agents are increasingly extended through Agent Skills, reusable artifacts that package natural-language metadata, procedural instructions, and execution-time resources for runtime use. As open-source skill marketplaces expand, users and agents increasingly rely on brief metadata to select third-party skills, making it difficult to detect inconsistencies between a skill's description and its true behavior, a problem we call cross-layer misalignment. To address this issue, we propose Progressive Loading-Aware Hierarchical Contrastive Learning (PL-HCL), an LLM-based framework that detects misalignment by modeling the layered structure of Agent Skills and learning cross-layer consistency. Using a normalized corpus of over 264,000 open-source skills and a human-verified challenge set, PL-HCL improves Macro-F1 from approximately 0.45 for unadapted baselines to 0.87-0.89 across evaluated LLM backbones. This approach offers an effective screening tool for users and operators, as well as design principles for detecting inconsistencies in layered digital artifacts.

cs.AI↗