Searcharxiv⌕ Search

arXiv subjects

Tao Yang

Publications and source records attributed to Tao Yang.

At least 109 records · Page 6Linked to original sources
arXiv2025

Time Resolution Characterization of 4H-SiC LGADs with a ${}^{90}$Sr Source

This work presents timing measurements of 4H-SiC Low Gain Avalanche Detectors (4H-SiC LGADs) using beta particles from a ${}^{90}$Sr source. The 4H-SiC LGADs exhibit fast signal responses, and a time resolution of 61~ps was achieved, comparable to that of standard Si LGADs. The present limitation in the time resolution of 4H-SiC LGADs appears to stem from limited charge generation. Nevertheless, their higher voltage tolerance and faster carrier drift suggest that, with increased charge collection, their timing performance could approach or even surpass that of Si LGADs. These results demonstrate the strong potential of 4H-SiC LGADs as a robust platform for precision timing in future 4D tracking detectors, while also highlighting that signal charge is the dominant factor currently limiting their performance, indicating that further optimization of gain and drift structures will be essential for future development.

physics.ins-det↗
arXiv2025

Distributed Constrained Online Nonconvex Optimization with Compressed Communication

This paper considers distributed online nonconvex optimization with time-varying inequality constraints over a network of agents. For a time-varying graph, we propose a distributed online primal-dual algorithm with compressed communication to efficiently utilize communication resources. We show that the proposed algorithm establishes an $\mathcal{O}( {{T^{\max \{ {1 - {θ_1},{θ_1}} \}}}} )$ network regret bound and an $\mathcal{O}( {T^{1 - {θ_1}/2}} )$ network cumulative constraint violation bound, where $T$ is the number of iterations and ${θ_1} \in ( {0,1} )$ is a user-defined trade-off parameter. When Slater's condition holds (i.e, there is a point that strictly satisfies the inequality constraints at all iterations), the network cumulative constraint violation bound is reduced to $\mathcal{O}( {T^{1 - {θ_1}}} )$. These bounds are comparable to the state-of-the-art results established by existing distributed online algorithms with perfect communication for distributed online convex optimization with (time-varying) inequality constraints. Finally, a simulation example is presented to validate the theoretical results.

math.OC↗
arXiv2025

Determination of ground states of one-dimensional quantum systems using the cluster iTEBD method

Within the framework of imaginary-time evolution for matrix product states, we introduce a cluster version of the infinite time-evolving block decimation algorithm for simulating quantum many-body systems, addressing the computational accuracy challenges in strongly correlated physics. By redefining the wave-function ansatz to incorporate multiple physical degrees of freedom, we enhance the representation of entanglement, thereby improving the accuracy of the ground states. Utilizing the Trotter-Suzuki decomposition and optimized truncation schemes, our method maintains roughly the same computational complexity while capturing more quantum correlations. We apply this approach to three nontrivial cases: the gapless spin-1/2 Heisenberg chain, the spin-1 anisotropic XXZD chain with a higher-order Gaussian-type phase transition, and a spin-1/2 twisted triangular prism hosting a magnetic plateau phase. Improved accuracy in physical quantities, such as magnetization, ground-state energy, and entanglement entropy, has been demonstrated. This method provides a scalable framework for studying complex quantum systems with high precision, making it suitable for situations where a pure increase in bond dimension alone cannot guarantee satisfactory results.

cond-mat.str-el↗
arXiv2025

Reinforcement Learning to Rank Using Coarse-grained Rewards

Learning to rank (LTR) plays a crucial role in various Information Retrieval (IR) tasks. Although supervised LTR methods based on fine-grained relevance labels (e.g., document-level annotations) have achieved significant success, their reliance on costly and potentially biased annotations limits scalability and alignment with realistic goals. In contrast, coarse-grained feedback signals, such as duration time and session-level engagement, are more accessible and affordable. Reinforcement Learning (RL) offers a promising framework to directly optimize these objectives using reward signals, but most existing Reinforcement Learning to Rank (RLTR) approaches suffer from high variance and low sample efficiency. Motivated by recent advances in large language models (LLMs), we re-examine the problem of RLTR with coarse-grained rewards and propose new RLTR methods based on widely used RL algorithms for LLMs. We systematically compare supervised learning and RL-based methods across various model architectures and coarse-grained reward functions on large-scale LTR benchmarks. Experimental results demonstrate that advanced RL methods can directly learn from coarse-grained rewards and outperform strong supervised learning baselines even with fine-grained labels. This shows the great potential of RLTR for metric-agnostic ranking optimization.

cs.IR↗
arXiv2025

WeChat-YATT: A Scalable, Simple, Efficient, and Production Ready Training Library

Reinforcement Learning from Human Feedback (RLHF) has emerged as a prominent paradigm for training large language models and multimodal systems. Despite the notable advances enabled by existing RLHF training frameworks, significant challenges remain to scale to complex multimodal workflows and adapt to dynamic workloads. In particular, current systems often encounter limitations related to controller scalability when managing large models, as well as inefficiencies in orchestrating intricate RLHF pipelines, especially in scenarios that require dynamic sampling and resource allocation. In this paper, we introduce WeChat-YATT Yet Another Transformer Trainer in WeChat, a simple, scalable, and balanced RLHF training framework specifically designed to address these challenges. WeChat-YATT features a parallel controller programming model that enables flexible and efficient orchestration of complex RLHF workflows, effectively mitigating bottlenecks associated with centralized controller architectures and facilitating scalability in large-scale data scenarios. In addition, we propose a dynamic placement schema that adaptively partitions computational resources and schedules workloads, thereby significantly reducing hardware idle time and improving GPU utilization under variable training conditions. We evaluate WeChat-YATT across diverse experimental scenarios, demonstrating its substantial throughput improvements over state-of-the-art RLHF training frameworks. Furthermore, WeChat-YATT has been successfully deployed to train models that support WeChat product features for a large-scale user base, underscoring its effectiveness and robustness in real-world applications. We have made WeChat-YATT publicly available at https://www.github.com/tencent/WeChat-YATT.

cs.LG↗
arXiv2025

The 1st International Workshop on Disentangled Representation Learning for Controllable Generation (DRL4Real): Methods and Results

This paper reviews the 1st International Workshop on Disentangled Representation Learning for Controllable Generation (DRL4Real), held in conjunction with ICCV 2025. The workshop aimed to bridge the gap between the theoretical promise of Disentangled Representation Learning (DRL) and its application in realistic scenarios, moving beyond synthetic benchmarks. DRL4Real focused on evaluating DRL methods in practical applications such as controllable generation, exploring advancements in model robustness, interpretability, and generalization. The workshop accepted 9 papers covering a broad range of topics, including the integration of novel inductive biases (e.g., language), the application of diffusion models to DRL, 3D-aware disentanglement, and the expansion of DRL into specialized domains like autonomous driving and EEG analysis. This summary details the workshop's objectives, the themes of the accepted papers, and provides an overview of the methodologies proposed by the authors.

cs.LG↗
arXiv2025

Dynamical Dark Energy Implies a Coupled Dark Sector: Insights from DESI DR2 via a Data-Driven Approach

Recent observations from the Dark Energy Spectroscopic Instrument (DESI) Data Release 2 (DR2) have revealed compelling evidence for dynamical dark energy, challenging the $Λ$CDM paradigm. In this work, we adopt a data-driven, model-independent approach to reconstruct the dark energy equation of state (EoS) and its potential interaction with dark matter using combined background cosmological datasets, including DESI DR2, cosmic chronometers, observational Hubble data, and Type Ia supernovae. Using Gaussian Process regression and a non-parametric formalism, we first confirm a $\sim 2σ$ indication of dynamical dark energy, featuring a phantom crossing around redshift $z \sim 0.4$, consistent with DESI results. We then explore the implications of dynamical EoS from DESI DR2 for dark sector coupling. Incorporating priors on the EoS from DESI DR2, we find a $\sim 2σ$ signal for non-zero interactions between dark energy and dark matter at low redshift. Our results suggest that if DESI's evidence for time-varying dark energy is confirmed, a coupled dark sector may be a necessary extension beyond $Λ$CDM.

astro-ph.CO↗
arXiv2025

G-Core: A Simple, Scalable and Balanced RLHF Trainer

Reinforcement Learning from Human Feedback (RLHF) has become an increasingly popular paradigm for training large language models (LLMs) and diffusion models. While existing RLHF training systems have enabled significant progress, they often face challenges in scaling to multi-modal and diffusion workflows and adapting to dynamic workloads. In particular, current approaches may encounter limitations in controller scalability, flexible resource placement, and efficient orchestration when handling complex RLHF pipelines, especially in scenarios involving dynamic sampling or generative reward modeling. In this paper, we present \textbf{G-Core}, a simple, scalable, and balanced RLHF training framework designed to address these challenges. G-Core introduces a parallel controller programming model, enabling flexible and efficient orchestration of complex RLHF workflows without the bottlenecks of a single centralized controller. Furthermore, we propose a dynamic placement schema that adaptively partitions resources and schedules workloads, significantly reducing hardware idle time and improving utilization, even under highly variable training conditions. G-Core has successfully trained models that support WeChat product features serving a large-scale user base, demonstrating its effectiveness and robustness in real-world scenarios. Our results show that G-Core advances the state of the art in RLHF training, providing a solid foundation for future research and deployment of large-scale, human-aligned models.

cs.LG↗
arXiv2025

Impact of Proton Irradiation on 4H-SiC Low Gain Avalanche Detectors (LGADs)

Silicon carbide (SiC) particle detectors have the potential to provide time resolutions and robust performance in extreme environments which exceed that of silicon detectors. In this work 4H-SiC low gain avalanche detectors (LGADs) and complementary PiN diodes were irradiated with 2.5 GeV protons at fluences up to 3.33$\times$10$^{14}$ p/cm$^2$. The electrostatic performance of both irradiated and non-irradiated devices was evaluated using current-voltage (I-V) and capacitance-voltage (C-V) measurements. Moreover, charge collection measurements using $α$ particles were also conducted. SiC LGADs displayed a loss in rectification and gain with increasing proton fluence. Additionally, the reduction in capacitance and OFF-state current pointed to compensation of the gain layer as a gain reducing mechanism. The introduction of radiation induced defects also hinders carrier acceleration reducing impact ionization, leading to further gain reduction. However, despite the reduction in device performance, the demonstration of a measurable signal and gain after irradiation points to the potential of SiC LGAD detectors for future high energy physics applications.

physics.ins-det↗
arXiv2025

UloRL:An Ultra-Long Output Reinforcement Learning Approach for Advancing Large Language Models' Reasoning Abilities

Recent advances in large language models (LLMs) have highlighted the potential of reinforcement learning with verifiable rewards (RLVR) to enhance reasoning capabilities through extended output sequences. However, traditional RL frameworks face inefficiencies when handling ultra-long outputs due to long-tail sequence distributions and entropy collapse during training. To address these challenges, we propose an Ultra-Long Output Reinforcement Learning (UloRL) approach for advancing large language models' reasoning abilities. Specifically, we divide ultra long output decoding into short segments, enabling efficient training by mitigating delays caused by long-tail samples. Additionally, we introduce dynamic masking of well-Mastered Positive Tokens (MPTs) to prevent entropy collapse. Experimental results demonstrate the effectiveness of our approach. On the Qwen3-30B-A3B model, RL with segment rollout achieved 2.06x increase in training speed, while RL training with 128k-token outputs improves the model's performance on AIME2025 from 70.9\% to 85.1\% and on BeyondAIME from 50.7\% to 61.9\%, even surpassing Qwen3-235B-A22B with remarkable gains. These findings underscore the potential of our methods to advance the reasoning capabilities of LLMs with ultra-long sequence generation. We will release our code and model for further use by the community.

cs.CL↗
arXiv2025

Distributed Optimization by Network Flows with Spatio-Temporal Compression

Several data compressors have been proposed in distributed optimization frameworks of network systems to reduce communication overhead in large-scale applications. In this paper, we demonstrate that effective information compression may occur over time or space during sequences of node communications in distributed algorithms, leading to the concept of spatio-temporal compressors. This abstraction classifies existing compressors and inspires new compressors as spatio-temporal compressors, with their effectiveness described by constructive stability criteria from nonlinear system theory. Subsequently, we incorporate these spatio-temporal compressors directly into standard continuous-time consensus flows and distributed primal-dual flows, establishing conditions ensuring exponential convergence. Additionally, we introduce a novel observer-based distributed primal-dual continuous flow integrated with spatio-temporal compressors, which provides broader convergence conditions. These continuous flows achieve exponential convergence to the global optimum when the objective function is strongly convex and can be discretized using Euler approximations. Finally, numerical simulations illustrate the versatility of the proposed spatio-temporal compressors and verify the convergence of

eess.SY↗
arXiv2025

Towards Solving More Challenging IMO Problems via Decoupled Reasoning and Proving

Automated Theorem Proving (ATP) in formal languages is a foundational challenge for AI. While Large Language Models (LLMs) have driven remarkable progress, a significant gap remains between their powerful informal reasoning capabilities and their weak formal proving performance. Recent studies show that the informal accuracy exceeds 80% while formal success remains below 8% on benchmarks like PutnamBench. We argue this gap persists because current state-of-the-art provers, by tightly coupling reasoning and proving, are trained with paradigms that inadvertently punish deep reasoning in favor of shallow, tactic-based strategies. To bridge this fundamental gap, we propose a novel framework that decouples high-level reasoning from low-level proof generation. Our approach utilizes two distinct, specialized models: a powerful, general-purpose Reasoner to generate diverse, strategic subgoal lemmas, and an efficient Prover to rigorously verify them. This modular design liberates the model's full reasoning potential and bypasses the pitfalls of end-to-end training. We evaluate our method on a challenging set of post-2000 IMO problems, a problem set on which no prior open-source prover has reported success. Our decoupled framework successfully solves 5 of these problems, demonstrating a significant step towards automated reasoning on exceptionally difficult mathematical challenges. To foster future research, we release our full dataset of generated and verified lemmas for a wide range of IMO problems, available at https://tencent-imo.github.io/ .

cs.LO↗
arXiv2025

Hunyuan-TurboS: Advancing Large Language Models through Mamba-Transformer Synergy and Adaptive Chain-of-Thought

As Large Language Models (LLMs) rapidly advance, we introduce Hunyuan-TurboS, a novel large hybrid Transformer-Mamba Mixture of Experts (MoE) model. It synergistically combines Mamba's long-sequence processing efficiency with Transformer's superior contextual understanding. Hunyuan-TurboS features an adaptive long-short chain-of-thought (CoT) mechanism, dynamically switching between rapid responses for simple queries and deep "thinking" modes for complex problems, optimizing computational resources. Architecturally, this 56B activated (560B total) parameter model employs 128 layers (Mamba2, Attention, FFN) with an innovative AMF/MF block pattern. Faster Mamba2 ensures linear complexity, Grouped-Query Attention minimizes KV cache, and FFNs use an MoE structure. Pre-trained on 16T high-quality tokens, it supports a 256K context length and is the first industry-deployed large-scale Mamba model. Our comprehensive post-training strategy enhances capabilities via Supervised Fine-Tuning (3M instructions), a novel Adaptive Long-short CoT Fusion method, Multi-round Deliberation Learning for iterative improvement, and a two-stage Large-scale Reinforcement Learning process targeting STEM and general instruction-following. Evaluations show strong performance: overall top 7 rank on LMSYS Chatbot Arena with a score of 1356, outperforming leading models like Gemini-2.0-Flash-001 (1352) and o4-mini-2025-04-16 (1345). TurboS also achieves an average of 77.9% across 23 automated benchmarks. Hunyuan-TurboS balances high performance and efficiency, offering substantial capabilities at lower inference costs than many reasoning models, establishing a new paradigm for efficient large-scale pre-trained models.

cs.CL↗
arXiv2025

Index Theorem and Vortex Kinetics in Bose-Einstein Condensates on a Haldane Sphere with a Magnetic Monopole

The geometry-gauge interplay constitutes a fundamental issue in quantum physics, with profound implications spanning from quantum gravity to topological matter. Here, we investigate the dynamic effects of geometry-gauge interplay in Bose-Einstein condensates (BECs) on a Haldane sphere with a magnetic monopole. We reveal an index theorem that establishes a correspondence between BEC vortices and the topology of the gauge field, enabling the construction of vortex-monopole composites. Furthermore, we derive the universal logarithmic interaction between composites, which governs the structure of the ground-state vortex lattice. By developing a kinetic theory, we predict scale-invariant vortex dynamics and an emergent duality. Both are confirmed through numerical simulations. This work first presents the dynamical coupling mechanism between spatial geometry and gauge fields, providing deep insights into superfluid systems with topological gauge structures in curved space.

cond-mat.quant-gas↗
arXiv2025

Seedance 1.0: Exploring the Boundaries of Video Generation Models

Notable breakthroughs in diffusion modeling have propelled rapid improvements in video generation, yet current foundational model still face critical challenges in simultaneously balancing prompt following, motion plausibility, and visual quality. In this report, we introduce Seedance 1.0, a high-performance and inference-efficient video foundation generation model that integrates several core technical improvements: (i) multi-source data curation augmented with precision and meaningful video captioning, enabling comprehensive learning across diverse scenarios; (ii) an efficient architecture design with proposed training paradigm, which allows for natively supporting multi-shot generation and jointly learning of both text-to-video and image-to-video tasks. (iii) carefully-optimized post-training approaches leveraging fine-grained supervised fine-tuning, and video-specific RLHF with multi-dimensional reward mechanisms for comprehensive performance improvements; (iv) excellent model acceleration achieving ~10x inference speedup through multi-stage distillation strategies and system-level optimizations. Seedance 1.0 can generate a 5-second video at 1080p resolution only with 41.4 seconds (NVIDIA-L20). Compared to state-of-the-art video generation models, Seedance 1.0 stands out with high-quality and fast video generation having superior spatiotemporal fluidity with structural stability, precise instruction adherence in complex multi-subject contexts, native multi-shot narrative coherence with consistent subject representation.

cs.CV↗
arXiv2025

Distributed satellite information networks: Architecture, enabling technologies, and trends

Driven by the vision of ubiquitous connectivity and wireless intelligence, the evolution of ultra-dense constellation-based satellite-integrated Internet is underway, now taking preliminary shape. Nevertheless, the entrenched institutional silos and limited, nonrenewable heterogeneous network resources leave current satellite systems struggling to accommodate the escalating demands of next-generation intelligent applications. In this context, the distributed satellite information networks (DSIN), exemplified by the cohesive clustered satellites system, have emerged as an innovative architecture, bridging information gaps across diverse satellite systems, such as communication, navigation, and remote sensing, and establishing a unified, open information network paradigm to support resilient space information services. This survey first provides a profound discussion about innovative network architectures of DSIN, encompassing distributed regenerative satellite network architecture, distributed satellite computing network architecture, and reconfigurable satellite formation flying, to enable flexible and scalable communication, computing and control. The DSIN faces challenges from network heterogeneity, unpredictable channel dynamics, sparse resources, and decentralized collaboration frameworks. To address these issues, a series of enabling technologies is identified, including channel modeling and estimation, cloud-native distributed MIMO cooperation, grant-free massive access, network routing, and the proper combination of all these diversity techniques. Furthermore, to heighten the overall resource efficiency, the cross-layer optimization techniques are further developed to meet upper-layer deterministic, adaptive and secure information services requirements. In addition, emerging research directions and new opportunities are highlighted on the way to achieving the DSIN vision.

cs.IT↗
arXiv2025

Indeterminate Probability Theory

Complex continuous or mixed joint distributions (e.g., P(Y | z_1, z_2, ..., z_N)) generally lack closed-form solutions, often necessitating approximations such as MCMC. This paper proposes Indeterminate Probability Theory (IPT), which makes the following contributions: (1) An observer-centered framework in which experimental outcomes are represented as distributions combining ground truth with observation error; (2) The introduction of three independence candidate axioms that enable a two-phase probabilistic inference framework; (3) The derivation of closed-form solutions for arbitrary complex joint distributions under this framework. Both the Indeterminate Probability Neural Network (IPNN) model and the non-neural multivariate time series forecasting application demonstrate IPT's effectiveness in modeling high-dimensional distributions, with successful validation up to 1000 dimensions. Importantly, IPT is consistent with classical probability theory and subsumes the frequentist equation in the limit of vanishing observation error.

cs.LG↗