arXiv subjects

Tao Song

Publications and source records attributed to Tao Song.

At least 19 recordsLinked to original sources

arXiv2026

GAUGE: Grading Agent-Built Financial Models Without a Golden Answer

Financial models combine public disclosures with analyst assumptions to produce forecasts and valuations. While some components can be checked mechanically, forecasts, discount rates, and target prices often admit multiple reasonable answers. Existing benchmarks nevertheless tend to grade such outputs against a single expert reference. Using independently built analyst models for the same companies, we find that across 108 directed pairs covering 65 companies, the median single-reference score is 0.33, 92.6% score below 0.70, and no same-vintage pair agrees on implied price within 10%. Point-tolerance grading can therefore penalize disagreement already present among professionals. We introduce GAUGE, a benchmark for evaluating agent-built valuation models against observed analyst practice rather than a single point answer. GAUGE uses 1,001 vendor-classified analyst workbooks and a 196-task evaluation set, with a three-layer observed-practice envelope, 56 auditable facets, eight validity gates, and deterministic structural checks. We validate the benchmark with a 55-participant known-groups study, company-grouped cross-fitting, and judge-stability audits. On the failure-aware score $\phi_0$, senior analysts average 88.3, juniors 66.0, and finance students 43.2. Across 24 agents and 1,011 scored generations, the best agent scores 53.4, above the student mean but below every senior and most juniors. It passes 93% of mechanical facets and 78% of judgment facets, with a fleet-median gap of 26 points. Current agents are substantially stronger at model construction than valuation judgment. We release the methodology, a gated de-identified data tier, a controlled training split, a versioned 48-task evaluation core, and a withheld refresh pool.

cs.LG↗

arXiv2026

WNOJ-LIO: A White-Noise-on-Jerk Motion-Prior EKF for High-Dynamic LiDAR-IMU Fusion

LiDAR-inertial odometry (LIO) is a key component of autonomous navigation, but high-dynamic driving exposes two coupled challenges: intra-scan motion distortion and vibration-contaminated inertial measurements. Most real-time LiDAR-inertial pipelines propagate the system state by integrating raw IMU measurements and then use the propagated trajectory for point cloud de-distortion, thereby propagating inertial noise into both the corrected scan and the subsequent scan-to-map registration. This paper presents WNOJ-LIO, a LiDAR-IMU fusion framework based on a White-Noise-on-Jerk (WNOJ) Extended Kalman Filter (EKF). WNOJ-LIO employs a decoupled WNOJ prior on $\R^3 \times \SO(3)$ for state prediction and treats the IMU as a high-frequency measurement source rather than the driver of state propagation. The resulting posterior state history is then used for LiDAR scan de-distortion and subsequent point-to-plane LiDAR updates. The decoupled process model enables closed-form covariance propagation, thereby bridging the gap between batch WNOJ Gaussian process (GP) trajectory priors and recursive filtering. Simulation results demonstrate improvements in acceleration and angular-velocity denoising, scan de-distortion, and localization accuracy over a FAST-LIO-style baseline. Real-world experiments were conducted using an autonomous racing car on four driving segments with maximum speeds ranging from 53 to 208~km/h, covering a wide range of vehicle vibration levels. The experiments further validate the proposed method and provide a comprehensive evaluation of its performance in estimating acceleration, angular velocity, body-frame linear velocity, attitude, and position under highly dynamic driving. The source code of WNOJ-LIO is publicly available at https://github.com/LvJohny/wnoj-ekf-lio.git.

cs.RO↗

arXiv2026

TianJi-Environ: An Autonomous AI Scientist for Atmospheric Environmental Research

As atmospheric environmental prediction continues to improve, interpretable validation of pollution mechanisms and feedback processes has become a main challenge in atmospheric chemistry. Yet mechanism validation based on complex numerical models still relies heavily on expert knowledge: mechanistic hypotheses must be operationalized into executable experiments, and model outputs must be organized into traceable evidence. We present TianJi-Environ, an auditable AI Scientist for atmospheric-chemistry mechanism validation. TianJi-Environ establishes the first WRF-Chem-based multi-agent framework that autonomously drives complex atmospheric-chemistry simulations, converting mechanistic hypotheses into executable configurations, testing experiments, and evidence criteria. Using ozone response and particulate-matter feedback as two representative examples, we demonstrate TianJi-Environ's capability for mechanism validation. In a summertime ozone case over the North China Plain, the system detects directionally consistent aerosol-radiation-interaction signals in shortwave radiation and boundary-layer height, but judges the evidence for ozone response to NOx control to be incomplete. In a wintertime PM2.5 case over the Guanzhong Basin, it localizes the unsupported link to insufficient propagation from black-carbon perturbation to particulate response and missing diagnostics of vertical absorptive heating. These results show that TianJi-Environ makes expert-driven mechanism validation explicit, structured, and auditable, offering a reproducible paradigm for multi-agent systems coupled with complex atmospheric-chemistry models.

physics.ao-ph↗

arXiv2026

Echoes within the Reasoning: Stealthy and Effective Watermarking via Chain of Thought

Large Language Models with Chain-of-Thought reasoning capabilities represent valuable intellectual property, yet existing black-box watermarking methods often trade robustness for reasoning fidelity by perturbing final answers or relying on fragile trigger patterns. We propose BiCoT, a watermarking framework that embeds ownership signals into the internal geometry of reasoning traces by aligning high-saliency structural anchors with a private signature subspace while regularizing ordinary control tokens to preserve semantic capacity. This design couples the watermark with reasoning-relevant representations, making removal difficult without disrupting the features that support coherent reasoning. To enable verification under model theft and representation drift, we introduce Robust Subspace Registration (RSR), a Top- logprob-based black-box verifier that uses sentinel tokens to calibrate systematic shifts in the output distribution. Experiments show that BiCoT preserves reasoning fidelity across diverse complex reasoning tasks while achieving robust detection under fine-tuning, quantization, model-level perturbations, and adaptive output-level attacks across in-domain and out-of-distribution settings.

cs.CR↗

arXiv2026

Volador 1.0: A Data-Driven Air-Sea Full-Coupling Regional Forecast Model with Submesoscale-Permitting Based on MOE-Swin-Transformer Framework

A data-driven air-sea full-coupling regional forecast model with submesoscale-permitting, named "Volador 1.0", is developed for the South China Sea (SCS). The model features a Swin-Transformer framework integrated with a Mixture-of-Experts (MoE) system, a latent space interaction architecture based on Cross-Grid Bidirectional Cross-Attention, and a fast-slow dual-branch architecture. Both the three-month hindcast test and the 15-day operational real-time forecasting demonstrate that Volador 1.0 has a very encouraging and promising performance in 0-72h forecasting of temperature and salinity in the 0-500m upper ocean as well as the sea surface height with root-mean-square-error (RMSE) or mean absolute error (MAE) smaller than or at least comparable to those from the reanalysis datasets REDOS V2.0 and GLORYS12 and the state-of-the-art regional numerical model Regional Ocean Modeling System (ROMS). In particular, Volador 1.0 demonstrates its capability of capturing/forecasting submesoscale processes including internal waves, with an energy spectrum well representing sub- to mesoscale energy cascade as expected by the classical turbulence theory. Further analysis based on ablation experiments shows that the air-sea full-coupling framework, which takes into account the dynamic exchanges of momentum and heat fluxes between the atmosphere and the ocean, indeed helps improve the model's performance compared to the non-full-coupling one. Volador 1.0, though still subject to refinement in the coming future with a large space for improvement, blazes a path for an accurate, fine and fast marine environment forecasting, and thus could help promote our capability of disaster prevention and mitigation in the SCS as well as in other coastal regions where these innovative techniques can be applied.

physics.ao-ph↗

arXiv2026

SCT-MOT: Enhancing Air-to-Air Multiple UAVs Tracking with Swarm-Coupled Motion and Trajectory Guidance

Air-to-air tracking of swarm UAVs presents significant challenges due to the complex nonlinear group motion and weak visual cues for small objects, which often cause detection failures, trajectory fragmentation, and identity switches. Although existing methods have attempted to improve performance by incorporating trajectory prediction, they model each object independently, neglecting the swarm-level motion dependencies. Their limited integration between motion prediction and appearance representation also weakens the spatio-temporal consistency required for tracking in visually ambiguous and cluttered environments, making it difficult to maintain coherent trajectories and reliable associations. To address these challenges, we propose SCT-MOT, a tracking framework that integrates Swarm-Coupled motion modeling and Trajectory-guided feature fusion. First, we develop a Swarm Motion-Aware Trajectory Prediction (SMTP) module jointly models historical trajectories and posture-aware appearance features from a swarm-level perspective, enabling more accurate forecasting of the nonlinear, coupled group trajectories. Second, we design a Trajectory-Guided Spatio-Temporal Feature Fusion (TG-STFF) module aligns predicted positions with historical visual cues and deeply integrates them with current frame features, enhancing temporal consistency and spatial discriminability for weak objects. Extensive experiments on three public air-to-air swarm UAV tracking datasets, including AIRMOT, MOT-FLY, and UAVSwarm, demonstrate that SMTP achieves more accurate trajectory forecasts and yields a 1.21\% IDF1 improvement over the state-of-the-art trajectory prediction module EqMotion when integrated into the same MOT framework. Overall, our SCT-MOT consistently achieves superior accuracy and robustness compared to state-of-the-art trackers across multiple metrics under complex swarm scenarios.

cs.CV↗

arXiv2026

TianJi:An autonomous AI meteorologist for discovering physical mechanisms in atmospheric science

Artificial intelligence (AI) has achieved breakthroughs comparable to traditional numerical models in data-driven weather forecasting, yet it remains essentially statistical fitting and struggles to uncover the physical causal mechanisms of the atmosphere. Physics-oriented mechanism research still heavily relies on domain knowledge and cumbersome engineering operations of human scientists, becoming a bottleneck restricting the efficiency of Earth system science exploration. Here, we propose TianJi - the first "AI meteorologist" system capable of autonomously driving complex numerical models to verify physical mechanisms. Powered by a large language model-driven multi-agent architecture, TianJi can autonomously conduct literature research and generate scientific hypotheses. We further decouple scientific research into cognitive planning and engineering execution: the meta-planner interprets hypotheses and devises experimental roadmaps, while a cohort of specialized worker agents collaboratively complete data preparation, model configuration, and multi-dimensional result analysis. In two classic atmospheric dynamic scenarios (squall-line cold pools and typhoon track deflections), TianJi accomplishes expert-level end-to-end experimental operations with zero human intervention, compressing the research cycle to a few hours. It also delivers detailed result analyses and autonomously judges and explains the validity of the hypotheses from outputs. TianJi reveals that the role of AI in Earth system science is transitioning from a "black-box predictor" to an "interpretable scientific collaborator", offering a new paradigm for high-throughput exploration of scientific mechanisms.

cs.AI↗

arXiv2026

FedMomentum: Preserving LoRA Training Momentum in Federated Fine-Tuning

Federated fine-tuning of large language models (LLMs) with low-rank adaptation (LoRA) offers a communication-efficient and privacy-preserving solution for task-specific adaptation. Naive aggregation of LoRA modules introduces noise due to mathematical incorrectness when averaging the downsampling and upsampling matrices independently. However, existing noise-free aggregation strategies inevitably compromise the structural expressiveness of LoRA, limiting its ability to retain client-specific adaptations by either improperly reconstructing the low-rank structure or excluding partially trainable components. We identify this problem as loss of training momentum, where LoRA updates fail to accumulate effectively across rounds, resulting in slower convergence and suboptimal performance. To address this, we propose FedMomentum, a novel framework that enables structured and momentum-preserving LoRA aggregation via singular value decomposition (SVD). Specifically, after aggregating low-rank updates in a mathematically correct manner, FedMomentum applies SVD to extract the dominant components that capture the main update directions. These components are used to reconstruct the LoRA modules with the same rank, while residual components can be retained and later merged into the backbone to preserve semantic information and ensure robustness. Extensive experiments across multiple tasks demonstrate that FedMomentum consistently outperforms prior state-of-the-art methods in convergence speed and final accuracy.

cs.LG↗

arXiv2026

FWeb3: A Practical Incentive-Aware Federated Learning Framework

Federated learning (FL) enables collaborative model training over distributed private data. However, sustaining open participation requires incentive mechanisms that compensate contributors for their resources and risks. Enabled by Web3 primitives, especially blockchains, recent FL proposals incorporate incentive mechanisms for open participation, yet most focus primarily on algorithmic design and overlook system-level challenges, including coordination efficiency, secure handling of model updates, and practical usability. We present FWeb3, a practical Web3-enabled FL framework for incentive-aware training in open environments. FWeb3 adopts a modular architecture that separates FL functions from Web3 support services, decoupling the off-chain training and data plane from on-chain settlement while preserving verifiable incentive execution. The framework supports pluggable aggregation and contribution evaluation methods and provides a browser-native DApp interface to lower the participation barrier. We evaluate FWeb3 in real-world settings and show that it supports end-to-end incentive-aware FL with transaction and data-transfer overheads of only 21.3% and 3.4% in WAN; FWeb3 also deploys from zero configuration in under 3 minutes and enables user onboarding in under 1 minute.

cs.DC↗

arXiv2026

SettleFL: Trustless and Scalable Reward Settlement Protocol for Federated Learning on Permissionless Blockchains (Extended version)

In open Federated Learning (FL) environments where no central authority exists, ensuring collaboration fairness relies on decentralized reward settlement, yet the prohibitive cost of permissionless blockchains directly clashes with the high-frequency, iterative nature of model training. Existing solutions either compromise decentralization or suffer from scalability bottlenecks due to linear on-chain costs. To address this, we present SettleFL, a trustless and scalable reward settlement protocol designed to minimize total economic friction by offering a family of two interoperable protocols. Leveraging a shared domain-specific circuit architecture, SettleFL offers two interoperable strategies: (1) a Commit-and-Challenge variant that minimizes on-chain costs via optimistic execution and dispute-driven arbitration, and (2) a Commit-with-Proof variant that guarantees instant finality through per-round validity proofs. This design allows the protocol to flexibly adapt to varying latency and cost constraints while enforcing rational robustness without trusted coordination. We conduct extensive experiments combining real FL workloads and controlled simulations. Results show that SettleFL remains practical when scaling to 800 participants, achieving substantially lower gas cost.

cs.CR↗

arXiv2026

pFedNavi: Structure-Aware Personalized Federated Vision-Language Navigation for Embodied AI

Vision-Language Navigation VLN requires large-scale trajectory instruction data from private indoor environments, raising significant privacy concerns. Federated Learning FL mitigates this by keeping data on-device, but vanilla FL struggles under VLNs' extreme cross-client heterogeneity in environments and instruction styles, making a single global model suboptimal. This paper proposes pFedNavi, a structure-aware and dynamically adaptive personalized federated learning framework tailored for VLN. Our key idea is to personalize where it matters: pFedNavi adaptively identifies client-specific layers via layer-wise mixing coefficients, and performs fine-grained parameter fusion on the selected components (e.g., the encoder-decoder projection and environment-sensitive decoder layers) to balance global knowledge sharing with local specialization. We evaluate pFedNavi on two standard VLN benchmarks, R2R and RxR, using both ResNet and CLIP visual representations. Across all metrics, pFedNavi consistently outperforms the FedAvg-based VLN baseline, achieving up to 7.5% improvement in navigation success rate and up to 7.8% gain in trajectory fidelity, while converging 1.38x faster under non-IID conditions.

cs.CV↗

arXiv2026

Efficient Data-Driven Production Scheduling in Pharmaceutical Manufacturing

This paper develops a data-driven, constraint-based optimization framework for a complex industrial job shop scheduling problem variant in pharmaceutical manufacturing. The formulation captures fixed routings and designated machines, explicit resource calendars with weekends and planned maintenance, and campaign sequencing through sequence-dependent cleaning times derived from site tables. The model is implemented with an open source constraint solver and evaluated on deterministic snapshots from a solid oral dosage facility under three objective formulations: makespan, makespan plus total tardiness, and makespan plus average tardiness. On three industrial instances of increasing size (10, 30, and 84 jobs) the proposed schedules dominate reference plans that solve a simplified variant without the added site rules. Makespan reductions reach $88.1\%$, $77.6\%$, and $54.9\%$ and total tardiness reductions reach $72.1\%$, $58.7\%$, and $18.2\%$, respectively. The composite objectives further decrease late job counts with negligible makespan change on the smaller instances and a modest increase on the largest instance. Optimality is proven on the small case, with relative gaps of $0.77\%$ and $14.92\%$ on the medium and large cases under a fixed time limit. The results show that a compact constraint programming formulation can deliver feasible, transparent schedules that respect site rules while improving adherence to due dates on real industrial data.

cs.PF↗

arXiv2026

CADET: Context-Conditioned Ads CTR Prediction With a Decoder-Only Transformer

Click-through rate (CTR) prediction is fundamental to online advertising systems. While Deep Learning Recommendation Models (DLRMs) with explicit feature interactions have long dominated this domain, recent advances in generative recommenders have shown promising results in content recommendation. However, adapting these transformer-based architectures to ads CTR prediction still presents unique challenges, including handling post-scoring contextual signals, maintaining offline-online consistency, and scaling to industrial workloads. We present CADET (Context-Conditioned Ads Decoder-Only Transformer), an end-to-end decoder-only transformer for ads CTR prediction deployed at LinkedIn. Our approach introduces several key innovations: (1) a context-conditioned decoding architecture with multi-tower prediction heads that explicitly model post-scoring signals such as ad position, resolving the chicken-and-egg problem between predicted CTR and ranking; (2) a self-gated attention mechanism that stabilizes training by adaptively regulating information flow at both representation and interaction levels; (3) a timestamp-based variant of Rotary Position Embedding (RoPE) that captures temporal relationships across timescales from seconds to months; (4) session masking strategies that prevent the model from learning dependencies on unavailable in-session events, addressing train-serve skew; and (5) production engineering techniques including tensor packing, sequence chunking, and custom Flash Attention kernels that enable efficient training and serving at scale. In online A/B testing, CADET achieves a 11.04\% CTR lift compared to the production LiRank baseline model, a hybrid ensemble of DCNv2 and sequential encoders. The system has been successfully deployed on LinkedIn's advertising platform, serving the main traffic for homefeed sponsored updates.

cs.LG↗

arXiv2025

Intelligent Multimodal Multi-Sensor Fusion-Based UAV Identification, Localization, and Countermeasures for Safeguarding Low-Altitude Economy

The development of the low-altitude economy has led to a growing prominence of uncrewed aerial vehicle (UAV) safety management issues. Therefore, accurate identification, real-time localization, and effective countermeasures have become core challenges in airspace security assurance. This paper introduces an integrated UAV management and control system based on deep learning, which integrates multimodal multi-sensor fusion perception, precise positioning, and collaborative countermeasures. By incorporating deep learning methods, the system combines radio frequency (RF) spectral feature analysis, radar detection, electro-optical identification, and other methods at the detection level to achieve the identification and classification of UAVs. At the localization level, the system relies on multi-sensor data fusion and the air-space-ground integrated communication network to conduct real-time tracking and prediction of UAV flight status, providing support for early warning and decision-making. At the countermeasure level, it adopts comprehensive measures that integrate ``soft kill'' and ``hard kill'', including technologies such as electromagnetic signal jamming, navigation spoofing, and physical interception, to form a closed-loop management and control process from early warning to final disposal, which significantly enhances the response efficiency and disposal accuracy of low-altitude UAV management.

eess.SP↗

arXiv2025

POLAR: Policy-based Layerwise Reinforcement Learning Method for Stealthy Backdoor Attacks in Federated Learning

Federated Learning (FL) enables decentralized model training across multiple clients without exposing local data, but its distributed feature makes it vulnerable to backdoor attacks. Despite early FL backdoor attacks modifying entire models, recent studies have explored the concept of backdoor-critical (BC) layers, which poison the chosen influential layers to maintain stealthiness while achieving high effectiveness. However, existing BC layers approaches rely on rule-based selection without consideration of the interrelations between layers, making them ineffective and prone to detection by advanced defenses. In this paper, we propose POLAR (POlicy-based LAyerwise Reinforcement learning), the first pipeline to creatively adopt RL to solve the BC layer selection problem in layer-wise backdoor attack. Different from other commonly used RL paradigm, POLAR is lightweight with Bernoulli sampling. POLAR dynamically learns an attack strategy, optimizing layer selection using policy gradient updates based on backdoor success rate (BSR) improvements. To ensure stealthiness, we introduce a regularization constraint that limits the number of modified layers by penalizing large attack footprints. Extensive experiments demonstrate that POLAR outperforms the latest attack methods by up to 40% against six state-of-the-art (SOTA) defenses.

cs.LG↗

arXiv2025

A Real-time Spatio-Temporal Trajectory Planner for Autonomous Vehicles with Semantic Graph Optimization

Planning a safe and feasible trajectory for autonomous vehicles in real-time by fully utilizing perceptual information in complex urban environments is challenging. In this paper, we propose a spatio-temporal trajectory planning method based on graph optimization. It efficiently extracts the multi-modal information of the perception module by constructing a semantic spatio-temporal map through separation processing of static and dynamic obstacles, and then quickly generates feasible trajectories via sparse graph optimization based on a semantic spatio-temporal hypergraph. Extensive experiments have proven that the proposed method can effectively handle complex urban public road scenarios and perform in real time. We will also release our codes to accommodate benchmarking for the research community

cs.RO↗

arXiv2025

Virtual Full-stack Scanning of Brain MRI via Imputing Any Quantised Code

Magnetic resonance imaging (MRI) is a powerful and versatile imaging technique, offering a wide spectrum of information about the anatomy by employing different acquisition modalities. However, in the clinical workflow, it is impractical to collect all relevant modalities due to the scan time and cost constraints. Virtual full-stack scanning aims to impute missing MRI modalities from available but incomplete acquisitions, offering a cost-efficient solution to enhance data completeness and clinical usability. Existing imputation methods often depend on global conditioning or modality-specific designs, which limit their generalisability across patient cohorts and imaging protocols. To address these limitations, we propose CodeBrain, a unified framework that reformulates various ``any-to-any'' imputation tasks as a region-level full-stack code prediction problem. CodeBrain adopts a two-stage pipeline: (1) it learns the compact representation of a complete MRI modality set by encoding it into scalar-quantised codes at the region level, enabling high-fidelity image reconstruction after decoding these codes along with modality-agnostic common features; (2) it trains a projection encoder to predict the full-stack code map from incomplete modalities via a grading-based design for diverse imputation scenarios. Extensive experiments on two public brain MRI datasets, i.e., IXI and BraTS 2023, demonstrate that CodeBrain consistently outperforms state-of-the-art methods, establishing a new benchmark for unified brain MRI imputation and enabling virtual full-stack scanning. Our code will be released at https://github.com/ycwu1997/CodeBrain.

cs.CV↗

arXiv2025

THOR: A Generic Energy Estimation Approach for On-Device Training

Battery-powered mobile devices (e.g., smartphones, AR/VR glasses, and various IoT devices) are increasingly being used for AI training due to their growing computational power and easy access to valuable, diverse, and real-time data. On-device training is highly energy-intensive, making accurate energy consumption estimation crucial for effective job scheduling and sustainable AI. However, the heterogeneity of devices and the complexity of models challenge the accuracy and generalizability of existing estimation methods. This paper proposes THOR, a generic approach for energy consumption estimation in deep neural network (DNN) training. First, we examine the layer-wise energy additivity property of DNNs and strategically partition the entire model into layers for fine-grained energy consumption profiling. Then, we fit Gaussian Process (GP) models to learn from layer-wise energy consumption measurements and estimate a DNN's overall energy consumption based on its layer-wise energy additivity property. We conduct extensive experiments with various types of models across different real-world platforms. The results demonstrate that THOR has effectively reduced the Mean Absolute Percentage Error (MAPE) by up to 30%. Moreover, THOR is applied in guiding energy-aware pruning, successfully reducing energy consumption by 50%, thereby further demonstrating its generality and potential.

cs.LG↗