arXiv subjects

Hao Zhang

Publications and source records attributed to Hao Zhang.

At least 469 records · Page 26Linked to original sources

arXiv2025

Scaling Up RL: Unlocking Diverse Reasoning in LLMs via Prolonged Training

Recent advancements in reasoning-focused language models such as OpenAI's O1 and DeepSeek-R1 have shown that scaling test-time computation-through chain-of-thought reasoning and iterative exploration-can yield substantial improvements on complex tasks like mathematics and code generation. These breakthroughs have been driven by large-scale reinforcement learning (RL), particularly when combined with verifiable reward signals that provide objective and grounded supervision. In this report, we investigate the effects of prolonged reinforcement learning on a small language model across a diverse set of reasoning domains. Our work identifies several key ingredients for effective training, including the use of verifiable reward tasks, enhancements to Group Relative Policy Optimization (GRPO), and practical techniques to improve training stability and generalization. We introduce controlled KL regularization, clipping ratio, and periodic reference policy resets as critical components for unlocking long-term performance gains. Our model achieves significant improvements over strong baselines, including +14.7% on math, +13.9% on coding, and +54.8% on logic puzzle tasks. To facilitate continued research, we release our model publicly.

cs.LG↗

arXiv2025

General Modular Harness for LLM Agents in Multi-Turn Gaming Environments

We introduce a modular harness design for LLM agents that composes of perception, memory, and reasoning components, enabling a single LLM or VLM backbone to tackle a wide spectrum of multi turn gaming environments without domain-specific engineering. Using classic and modern game suites as low-barrier, high-diversity testbeds, our framework provides a unified workflow for analyzing how each module affects performance across dynamic interactive settings. Extensive experiments demonstrate that the harness lifts gameplay performance consistently over un-harnessed baselines and reveals distinct contribution patterns, for example, memory dominates in long-horizon puzzles while perception is critical in vision noisy arcades. These findings highlight the effectiveness of our modular harness design in advancing general-purpose agent, given the familiarity and ubiquity of games in everyday human experience.

cs.AI↗

arXiv2025

Adaptive Tool Use in Large Language Models with Meta-Cognition Trigger

Large language models (LLMs) have shown remarkable emergent capabilities, transforming the execution of functional tasks by leveraging external tools for complex problems that require specialized processing or up-to-date data. While existing research expands LLMs access to diverse tools (e.g., program interpreters, search engines, calculators), the necessity of using these tools is often overlooked, leading to indiscriminate tool invocation. This naive approach raises two key issues: increased latency due to unnecessary tool calls, and potential errors resulting from faulty interactions with external tools. In this paper, we introduce meta-cognition as a proxy for LLMs self-assessment of their capabilities, reflecting the model's awareness of its own limitations. Based on this, we propose MeCo, an adaptive decision-making strategy for external tool use. MeCo quantifies metacognitive scores by capturing high-level cognitive signals in the representation space, guiding when to invoke tools. Notably, MeCo is fine-tuning-free and incurs minimal cost. Experiments across multiple backbone models and benchmarks show that MeCo reliably detects LLMs' internal cognitive signals and significantly improves tool-use decision-making.

cs.CL↗

arXiv2025

A Federated Learning-based Lightweight Network with Zero Trust for UAV Authentication

Unmanned aerial vehicles (UAVs) are increasingly being integrated into next-generation networks to enhance communication coverage and network capacity. However, the dynamic and mobile nature of UAVs poses significant security challenges, including jamming, eavesdropping, and cyber-attacks. To address these security challenges, this paper proposes a federated learning-based lightweight network with zero trust for enhancing the security of UAV networks. A novel lightweight spectrogram network is proposed for UAV authentication and rejection, which can effectively authenticate and reject UAVs based on spectrograms. Experiments highlight LSNet's superior performance in identifying both known and unknown UAV classes, demonstrating significant improvements over existing benchmarks in terms of accuracy, model compactness, and storage requirements. Notably, LSNet achieves an accuracy of over $80\%$ for known UAV types and an Area Under the Receiver Operating Characteristic (AUROC) of $0.7$ for unknown types when trained with all five clients. Further analyses explore the impact of varying the number of clients and the presence of unknown UAVs, reinforcing the practical applicability and effectiveness of our proposed framework in real-world FL scenarios.

eess.SP↗

arXiv2025

Visualization of nonlinear optics in a microresonator

A precise understanding of nonlinear optical phenomena in whispering gallery mode (WGM) microresonators is crucial for developing next-generation integrated photonic devices. Applications include on-chip sensors for biomedical use, optical memories for all-optical networks and frequency combs for optical clocks. However, our ability to spatially localize nonlinear optical processes within microresonators has been limited because optical feedback is often only collected through a bus waveguide. In this study, we present the direct visualization of nonlinear optical processes using scattering patterns captured by a short-wave infrared (SWIR) camera. Through systematic analysis of these scattering patterns, we can distinguish between different nonlinear effects occurring within the microresonator. Direct imaging of nonlinear processes in microresonators can significantly impact many applications, including the optimization of soliton frequency combs, real-time debugging of photonic circuits, microresonator-based memories, and chip-based data switching in telecom circuits.

physics.optics↗

arXiv2025

Fast-dLLM: Training-free Acceleration of Diffusion LLM by Enabling KV Cache and Parallel Decoding

Diffusion-based large language models (Diffusion LLMs) have shown promise for non-autoregressive text generation with parallel decoding capabilities. However, the practical inference speed of open-sourced Diffusion LLMs often lags behind autoregressive models due to the lack of Key-Value (KV) Cache and quality degradation when decoding multiple tokens simultaneously. To bridge this gap, we introduce a novel block-wise approximate KV Cache mechanism tailored for bidirectional diffusion models, enabling cache reuse with negligible performance drop. Additionally, we identify the root cause of generation quality degradation in parallel decoding as the disruption of token dependencies under the conditional independence assumption. To address this, we propose a confidence-aware parallel decoding strategy that selectively decodes tokens exceeding a confidence threshold, mitigating dependency violations and maintaining generation quality. Experimental results on LLaDA and Dream models across multiple LLM benchmarks demonstrate up to \textbf{27.6$\times$ throughput} improvement with minimal accuracy loss, closing the performance gap with autoregressive models and paving the way for practical deployment of Diffusion LLMs.

cs.CL↗

arXiv2025

RapidStore: An Efficient Dynamic Graph Storage System for Concurrent Queries

Dynamic graph storage systems are essential for real-time applications such as social networks and recommendation, where graph data continuously evolves. However, they face significant challenges in efficiently handling concurrent read and write operations. We find that existing methods suffer from write queries interfering with read efficiency, substantial time and space overhead due to per-edge versioning, and an inability to balance performance, such as slow searches under concurrent workloads. To address these issues, we propose RapidStore, a holistic approach for efficient in-memory dynamic graph storage designed for read-intensive workloads. Our key idea is to exploit the characteristics of graph queries through a decoupled system design that separates the management of read and write queries and decouples version data from graph data. Particularly, we design an efficient dynamic graph store to cooperate with the graph concurrency control mechanism. Experimental results demonstrate that RapidStore enables fast and scalable concurrent graph queries, effectively balancing the performance of inserts, searches, and scans, and significantly improving efficiency in dynamic graph storage systems.

cs.DB↗

arXiv2025

RCA: Region Conditioned Adaptation for Visual Abductive Reasoning

Visual abductive reasoning aims to make likely explanations for visual observations. We propose a simple yet effective Region Conditioned Adaptation, a hybrid parameter-efficient fine-tuning method that equips the frozen CLIP with the ability to infer explanations from local visual cues. We encode "local hints" and "global contexts" into visual prompts of the CLIP model separately at fine and coarse-grained levels. Adapters are used for fine-tuning CLIP models for downstream tasks and we design a new attention adapter, that directly steers the focus of the attention map with trainable query and key projections of a frozen CLIP model. Finally, we train our new model with a modified contrastive loss to regress the visual feature simultaneously toward features of literal description and plausible explanations. The loss enables CLIP to maintain both perception and reasoning abilities. Experiments on the Sherlock visual abductive reasoning benchmark show that the RCA significantly outstands previous SOTAs, ranking the 1st on the leaderboards (e.g., Human Acc: RCA 31.74 $\textit{vs}$ CPT-CLIP 29.58, higher =better). We also validate the RCA is generalizable to local perception benchmarks like RefCOCO. We open-source our project at https://github.com/LUNAProject22/RPA.

cs.CV↗

arXiv2025

Automated Vehicles Should be Connected with Natural Language

Multi-agent collaborative driving promises improvements in traffic safety and efficiency through collective perception and decision making. However, existing communication media -- including raw sensor data, neural network features, and perception results -- suffer limitations in bandwidth efficiency, information completeness, and agent interoperability. Moreover, traditional approaches have largely ignored decision-level fusion, neglecting critical dimensions of collaborative driving. In this paper we argue that addressing these challenges requires a transition from purely perception-oriented data exchanges to explicit intent and reasoning communication using natural language. Natural language balances semantic density and communication bandwidth, adapts flexibly to real-time conditions, and bridges heterogeneous agent platforms. By enabling the direct communication of intentions, rationales, and decisions, it transforms collaborative driving from reactive perception-data sharing into proactive coordination, advancing safety, efficiency, and transparency in intelligent transportation systems.

cs.MA↗

arXiv2025

High-Precision Dichotomous Image Segmentation via Probing Diffusion Capacity

In the realm of high-resolution (HR), fine-grained image segmentation, the primary challenge is balancing broad contextual awareness with the precision required for detailed object delineation, capturing intricate details and the finest edges of objects. Diffusion models, trained on vast datasets comprising billions of image-text pairs, such as SD V2.1, have revolutionized text-to-image synthesis by delivering exceptional quality, fine detail resolution, and strong contextual awareness, making them an attractive solution for high-resolution image segmentation. To this end, we propose DiffDIS, a diffusion-driven segmentation model that taps into the potential of the pre-trained U-Net within diffusion models, specifically designed for high-resolution, fine-grained object segmentation. By leveraging the robust generalization capabilities and rich, versatile image representation prior of the SD models, coupled with a task-specific stable one-step denoising approach, we significantly reduce the inference time while preserving high-fidelity, detailed generation. Additionally, we introduce an auxiliary edge generation task to not only enhance the preservation of fine details of the object boundaries, but reconcile the probabilistic nature of diffusion with the deterministic demands of segmentation. With these refined strategies in place, DiffDIS serves as a rapid object mask generation model, specifically optimized for generating detailed binary maps at high resolutions, while demonstrating impressive accuracy and swift processing. Experiments on the DIS5K dataset demonstrate the superiority of DiffDIS, achieving state-of-the-art results through a streamlined inference process. The source code will be publicly available at https://github.com/qianyu-dlut/DiffDIS.

cs.CV↗

arXiv2025

Spectroastrometry and Reverberation Mapping of Active Galactic Nuclei. II. Measuring Geometric Distances and Black Hole Masses of Four Nearby Quasars

The geometric distances of active galactic nuclei (AGNs) are challenging to measure because of their exceptionally compact structure yet vast cosmic distances. A combination of spectroastrometry and reverberation mapping (SARM) of broad-line regions (BLRs) constitutes a novel means to probe the geometric distance of AGNs, which has recently become practically feasible owing to successful interferometric observations with VLTI/GRAVITY. Here, we perform SARM analysis of four nearby quasars: Mrk 509, PDS 456, 3C 273, and NGC 3783. Results for the former two are reported for the first time and the latter two are revisited using our improved BLR dynamical modeling that includes the radial-dependent responsivity of BLRs. This allows us to self-consistently account for the emissivity weighting of the BLR in spectroastrometry and responsivity weighting in reverberation mapping. We obtain angular-diameter distances of the four quasars, from which we derive a Hubble constant of $H_0=69_{-10}^{+12}\,\rm km\,s^{-1}\,Mpc^{-1}$. Although this constitutes a large uncertainty for a measurement of $H_0$, it is anticipated that the precision will improve to a competitive level once a greater number of AGNs are accessible following the upgrade of GRAVITY in the near future. From SARM analysis, the black hole masses of the four quasars are also measured with the statistical uncertainty ranging from 0.06 to 0.23 dex, consistent with the correlations between black hole masses and properties of the host bulges.

astro-ph.GA↗

arXiv2025

PhysRig: Differentiable Physics-Based Skinning and Rigging Framework for Realistic Articulated Object Modeling

Skinning and rigging are fundamental components in animation, articulated object reconstruction, motion transfer, and 4D generation. Existing approaches predominantly rely on Linear Blend Skinning (LBS), due to its simplicity and differentiability. However, LBS introduces artifacts such as volume loss and unnatural deformations, and it fails to model elastic materials like soft tissues, fur, and flexible appendages (e.g., elephant trunks, ears, and fatty tissues). In this work, we propose PhysRig: a differentiable physics-based skinning and rigging framework that overcomes these limitations by embedding the rigid skeleton into a volumetric representation (e.g., a tetrahedral mesh), which is simulated as a deformable soft-body structure driven by the animated skeleton. Our method leverages continuum mechanics and discretizes the object as particles embedded in an Eulerian background grid to ensure differentiability with respect to both material properties and skeletal motion. Additionally, we introduce material prototypes, significantly reducing the learning space while maintaining high expressiveness. To evaluate our framework, we construct a comprehensive synthetic dataset using meshes from Objaverse, The Amazing Animals Zoo, and MixaMo, covering diverse object categories and motion patterns. Our method consistently outperforms traditional LBS-based approaches, generating more realistic and physically plausible results. Furthermore, we demonstrate the applicability of our framework in the pose transfer task highlighting its versatility for articulated object modeling.

cs.CV↗

arXiv2025

Synergy of Rashba and Topological Effects for High-Performance Bismuth-Based Thermoelectrics

Band convergence is a key strategy for enhancing thermoelectric (TE) performance. Herein, we demonstrate a promising approach to enhance band convergence through inducing Rashba splitting in topological insulators. Theoretically designed Janus $β$-Bi$_2$Se$_2$Te and $β$-Bi$_2$Te$_2$Se exhibited inherent topological band inversion and Rashba splitting due to the strong spin-orbit coupling (SOC) with broken inversion symmetry. These characteristics synergistically improve band convergence, leading to a substantially enhanced power factor. Meanwhile, the Janus structural asymmetry suppresses lattice thermal conductivity. Consequently, Janus structures achieve boosted TE performance, especially for $β$-Bi$_2$Se$_2$Te, peaking figure of merit ($zT$) of 2.82. This work establishes a new framework for designing Janus topological compounds with high TE performance by the synergistic effect of Rashba splitting and band inversion.

cond-mat.mtrl-sci↗

arXiv2025

MagicPose4D: Crafting Articulated Models with Appearance and Motion Control

With the success of 2D and 3D visual generative models, there is growing interest in generating 4D content. Existing methods primarily rely on text prompts to produce 4D content, but they often fall short of accurately defining complex or rare motions. To address this limitation, we propose MagicPose4D, a novel framework for refined control over both appearance and motion in 4D generation. Unlike current 4D generation methods, MagicPose4D accepts monocular videos or mesh sequences as motion prompts, enabling precise and customizable motion control. MagicPose4D comprises two key modules: (i) Dual-Phase 4D Reconstruction Module, which operates in two phases. The first phase focuses on capturing the model's shape using accurate 2D supervision and less accurate but geometrically informative 3D pseudo-supervision without imposing skeleton constraints. The second phase extracts the 3D motion (skeleton poses) using more accurate pseudo-3D supervision, obtained in the first phase and introduces kinematic chain-based skeleton constraints to ensure physical plausibility. Additionally, we propose a Global-local Chamfer loss that aligns the overall distribution of predicted mesh vertices with the supervision while maintaining part-level alignment without extra annotations. (ii) Cross-category Motion Transfer Module, which leverages the extracted motion from the 4D reconstruction module and uses a kinematic-chain-based skeleton to achieve cross-category motion transfer. It ensures smooth transitions between frames through dynamic rigidity, facilitating robust generalization without additional training. Through extensive experiments, we demonstrate that MagicPose4D significantly improves the accuracy and consistency of 4D content generation, outperforming existing methods in various benchmarks.

cs.CV↗

arXiv2025

Learning Adaptive Lighting via Channel-Aware Guidance

Learning lighting adaptation is a crucial step in achieving good visual perception and supporting downstream vision tasks. Current research often addresses individual light-related challenges, such as high dynamic range imaging and exposure correction, in isolation. However, we identify shared fundamental properties across these tasks: i) different color channels have different light properties, and ii) the channel differences reflected in the spatial and frequency domains are different. Leveraging these insights, we introduce the channel-aware Learning Adaptive Lighting Network (LALNet), a multi-task framework designed to handle multiple light-related tasks efficiently. Specifically, LALNet incorporates color-separated features that highlight the unique light properties of each color channel, integrated with traditional color-mixed features by Light Guided Attention (LGA). The LGA utilizes color-separated features to guide color-mixed features focusing on channel differences and ensuring visual consistency across all channels. Additionally, LALNet employs dual domain channel modulation for generating color-separated features and a mixed channel modulation and light state space module for producing color-mixed features. Extensive experiments on four representative light-related tasks demonstrate that LALNet significantly outperforms state-of-the-art methods on benchmark tests and requires fewer computational resources. We provide an anonymous online demo at https://xxxxxx2025.github.io/LALNet/.

cs.CV↗

arXiv2025

Analyzing LLMs' Knowledge Boundary Cognition Across Languages Through the Lens of Internal Representations

While understanding the knowledge boundaries of LLMs is crucial to prevent hallucination, research on the knowledge boundaries of LLMs has predominantly focused on English. In this work, we present the first study to analyze how LLMs recognize knowledge boundaries across different languages by probing their internal representations when processing known and unknown questions in multiple languages. Our empirical studies reveal three key findings: 1) LLMs' perceptions of knowledge boundaries are encoded in the middle to middle-upper layers across different languages. 2) Language differences in knowledge boundary perception follow a linear structure, which motivates our proposal of a training-free alignment method that effectively transfers knowledge boundary perception ability across languages, thereby helping reduce hallucination risk in low-resource languages; 3) Fine-tuning on bilingual question pair translation further enhances LLMs' recognition of knowledge boundaries across languages. Given the absence of standard testbeds for cross-lingual knowledge boundary analysis, we construct a multilingual evaluation suite comprising three representative types of knowledge boundary data. Our code and datasets are publicly available at https://github.com/DAMO-NLP-SG/LLM-Multilingual-Knowledge-Boundaries.

cs.CL↗

arXiv2025

Scaling Speculative Decoding with Lookahead Reasoning

Reasoning models excel by generating long chain-of-thoughts, but decoding the resulting thousands of tokens is slow. Token-level speculative decoding (SD) helps, but its benefit is capped, because the chance that an entire $γ$-token guess is correct falls exponentially as $γ$ grows. This means allocating more compute for longer token drafts faces an algorithmic ceiling -- making the speedup modest and hardware-agnostic. We raise this ceiling with Lookahead Reasoning, which exploits a second, step-level layer of parallelism. Our key insight is that reasoning models generate step-by-step, and each step needs only to be semantically correct, not exact token matching. In Lookahead Reasoning, a lightweight draft model proposes several future steps; the target model expands each proposal in one batched pass, and a verifier keeps semantically correct steps while letting the target regenerate any that fail. Token-level SD still operates within each reasoning step, so the two layers of parallelism multiply. We show Lookahead Reasoning lifts the peak speedup of SD both theoretically and empirically. Across GSM8K, AIME, and other benchmarks, Lookahead Reasoning improves the speedup of SD from 1.4x to 2.1x while preserving answer quality, and its speedup scales better with additional GPU throughput. Our code is available at https://github.com/hao-ai-lab/LookaheadReasoning

cs.LG↗

arXiv2025

Kimi-VL Technical Report

We present Kimi-VL, an efficient open-source Mixture-of-Experts (MoE) vision-language model (VLM) that offers advanced multimodal reasoning, long-context understanding, and strong agent capabilities - all while activating only 2.8B parameters in its language decoder (Kimi-VL-A3B). Kimi-VL demonstrates strong performance across challenging domains: as a general-purpose VLM, Kimi-VL excels in multi-turn agent tasks (e.g., OSWorld), matching flagship models. Furthermore, it exhibits remarkable capabilities across diverse challenging vision language tasks, including college-level image and video comprehension, OCR, mathematical reasoning, and multi-image understanding. In comparative evaluations, it effectively competes with cutting-edge efficient VLMs such as GPT-4o-mini, Qwen2.5-VL-7B, and Gemma-3-12B-IT, while surpassing GPT-4o in several key domains. Kimi-VL also advances in processing long contexts and perceiving clearly. With a 128K extended context window, Kimi-VL can process diverse long inputs, achieving impressive scores of 64.5 on LongVideoBench and 35.1 on MMLongBench-Doc. Its native-resolution vision encoder, MoonViT, further allows it to see and understand ultra-high-resolution visual inputs, achieving 83.2 on InfoVQA and 34.5 on ScreenSpot-Pro, while maintaining lower computational cost for common tasks. Building upon Kimi-VL, we introduce an advanced long-thinking variant: Kimi-VL-Thinking-2506. Developed through long chain-of-thought (CoT) supervised fine-tuning (SFT) and reinforcement learning (RL), the latest model exhibits strong long-horizon reasoning capabilities (64.0 on MMMU, 46.3 on MMMU-Pro, 56.9 on MathVision, 80.1 on MathVista, 65.2 on VideoMMMU) while obtaining robust general abilities. Code and models are publicly accessible at https://github.com/MoonshotAI/Kimi-VL.

cs.CV↗