arXiv subjects

Libo Zhang

Publications and source records attributed to Libo Zhang.

At least 37 records · Page 2Linked to original sources

arXiv2026

Sparrow: Text-Anchored Window Attention with Visual-Semantic Glimpsing for Speculative Decoding in Video LLMs

Although speculative decoding is widely used to accelerate Vision-Language Models (VLMs) inference, it faces severe performance collapse when applied to Video Large Language Models (Vid-LLMs). The draft model typically falls into the trap of attention dilution and negative visual gain due to key-value cache explosion and context window mismatches. We observe a visual semantic internalization phenomenon in Vid-LLMs, indicating that critical visual semantics are implicitly encoded into text hidden states during deep-layer interactions, which renders raw visual inputs structurally redundant during deep inference. To address this, we propose the Sparrow framework, which first utilizes visually-aware text-anchored window attention via hidden state reuse to fully offload visual computation to the target model, and leverages intermediate-layer visual state bridging to train the draft model with semantic-rich intermediate states, thereby filtering out low-level visual noise. Additionally, a multi-token prediction strategy is introduced to bridge the training-inference distribution shift. Experiments show that Sparrow achieves an average speedup of 2.82x even with 25k visual tokens, effectively resolving the performance degradation in long sequences and offering a practical solution for real-time long video tasks.

cs.CV↗

arXiv2026

Quantum dynamics of microwave photons in synthetic frequency dimension

Synthetic frequency dimension offers a powerful approach to simulate lattice models and control photon dynamics. However, extending this concept into the quantum regime, particularly at the single-photon level, has remained challenging in photonic platforms. Here, we demonstrate quantum-state initialization and detection of single-photon evolutions within a synthetic frequency lattice by integrating a superconducting qubit with a 16-meter aluminum coaxial cable. A tunable superconducting quantum interference device (SQUID)-based modulator is employed to synthesize lattice couplings and artificial gauge fields. We observe single-photon quantum random walks and Bloch oscillations, as well as nonadiabatic, unidirectional frequency conversion under rapid temporal modulation of the lattice Hamiltonian, together with band-structure measurements. The lattice connectivity can be readily reconfigured to construct higher-dimensional lattices using multiple drive tones. Our results establish superconducting quantum circuits as a versatile platform for programmable Hamiltonians and extensible synthetic lattices with flexible single-photon control.

quant-ph↗

arXiv2026

BYOS: Knowledge-driven Large Language Models Bring Your Own Operating System More Excellent

Operating system (OS) kernel tuning is a critical yet challenging problem for performance optimization, due to the large configuration space, complex interdependencies among configuration options, and the rapid evolution of kernel versions. Recent work has explored large language models (LLMs) for automated kernel tuning, but existing approaches often suffer from hallucinated configurations, limited interpretability, and poor robustness across workloads and kernel versions. We propose BYOS, a knowledge-driven framework that grounds LLM-based Linux kernel tuning in structured domain knowledge. BYOS incorporates three key components: (1) structured knowledge construction and mapping to bridge the semantic gap, (2) knowledge-driven configuration generation to refine the search space, and (3) continuous knowledge maintenance to adapt to kernel evolution. We evaluate BYOS on diverse workloads across multiple Linux distributions and kernel versions. Experimental results show that BYOS consistently outperforms state-of-the-art tuning baselines, achieving 7.1%-155.4% performance improvement while substantially reducing invalid configurations. These results demonstrate the effectiveness of integrating structured knowledge with LLMs for robust and scalable system optimization. The code of BYOS is available at https://github.com/LHY-24/BYOS.

cs.OS↗

arXiv2026

VisRefiner: Learning from Visual Differences for Screenshot-to-Code Generation

Screenshot-to-code generation aims to translate user interface screenshots into executable frontend code that faithfully reproduces the target layout and style. Existing multimodal large language models perform this mapping directly from screenshots but are trained without observing the visual outcomes of their generated code. In contrast, human developers iteratively render their implementation, compare it with the design, and learn how visual differences relate to code changes. Inspired by this process, we propose VisRefiner, a training framework that enables models to learn from visual differences between rendered predictions and reference designs. We construct difference-aligned supervision that associates visual discrepancies with corresponding code edits, allowing the model to understand how appearance variations arise from implementation changes. Building on this, we introduce a reinforcement learning stage for self-refinement, where the model improves its generated code by observing both the rendered output and the target design, identifying their visual differences, and updating the code accordingly. Experiments show that VisRefiner substantially improves single-step generation quality and layout fidelity, while also endowing models with strong self-refinement ability. These results demonstrate the effectiveness of learning from visual differences for advancing screenshot-to-code generation.

cs.CV↗

arXiv2026

Quantum-enhanced metrology with large Fock states

Quantum metrology uses non-classical states, such as Fock states with a specific number of photons, to achieve an advantage over classical sensing methods. Typically, quantum metrological performance can be enhanced by increasing the involved excitation numbers, for example, by using large photon-number Fock states. However, manipulating these states and demonstrating a quantum metrological advantage is experimentally challenging. Here we present an efficient method for generating large Fock states approaching 100 photons within a superconducting microwave cavity through the development of a programmable photon number filter. Using these states in displacement and phase measurements, we demonstrate quantum-enhanced metrology approaching the Heisenberg scaling for 40-photon Fock states and achieve a maximum metrological gain of up to 14.8 dB, highlighting the metrological advantages of large Fock states. Our study could be readily extended to mechanical and optical systems, promising potential applications in weak force detection and dark matter searches.

quant-ph↗

arXiv2026

A biased-erasure cavity qubit with hardware-efficient quantum error detection

Erasure qubits are beneficial for quantum error correction due to their relaxed threshold requirements. While dual-rail erasure qubits have been demonstrated with a strong error hierarchy in circuit quantum electrodynamics, biased-erasure qubits -- where erasures originate predominantly from one logical basis state -- offer further advantages. Here, we realize a hardware-efficient biased-erasure qubit encoded in the vacuum and two-photon Fock states of a single microwave cavity. The qubit exhibits an erasure bias ratio of over 265. By using a transmon ancilla for logical measurements and mid-circuit erasure detections, we achieve logical state assignment errors below 1% and convert over 99.3% leakage errors into detected erasures. After postselection against erasures, we achieve effective logical relaxation and dephasing rates of $(6.2~\mathrm{ms})^{-1}$ and $(3.1~\mathrm{ms})^{-1}$, respectively, which exceed the erasure error rate by factors of 31 and 15, establishing a strong error hierarchy within the logical subspace. These postselected error rates indicate a coherence gain of about 6.0 beyond the break-even point set by the best physical qubit encoded in the two lowest Fock states in the cavity. Moreover, randomized benchmarking with interleaved erasure detections reveals a residual logical gate error of 0.29%. This work establishes a compact and hardware-efficient platform for biased-erasure qubits, promising concatenations into outer-level stabilizer codes toward fault-tolerant quantum computation.

quant-ph↗

arXiv2026

EgoReAct: Egocentric Video-Driven 3D Human Reaction Generation

Humans exhibit adaptive, context-sensitive responses to egocentric visual input. However, faithfully modeling such reactions from egocentric video remains challenging due to the dual requirements of strictly causal generation and precise 3D spatial alignment. To tackle this problem, we first construct the Human Reaction Dataset (HRD) to address data scarcity and misalignment by building a spatially aligned egocentric video-reaction dataset, as existing datasets (e.g., ViMo) suffer from significant spatial inconsistency between the egocentric video and reaction motion, e.g., dynamically moving motions are always paired with fixed-camera videos. Leveraging HRD, we present EgoReAct, the first autoregressive framework that generates 3D-aligned human reaction motions from egocentric video streams in real-time. We first compress the reaction motion into a compact yet expressive latent space via a Vector Quantised-Variational AutoEncoder and then train a Generative Pre-trained Transformer for reaction generation from the visual input. EgoReAct incorporates 3D dynamic features, i.e., metric depth, and head dynamics during the generation, which effectively enhance spatial grounding. Extensive experiments demonstrate that EgoReAct achieves remarkably higher realism, spatial consistency, and generation efficiency compared with prior methods, while maintaining strict causality during generation. We will release code, models, and data upon acceptance.

cs.CV↗

arXiv2025

Rhea: Role-aware Heuristic Episodic Attention for Conversational LLMs

Large Language Models (LLMs) have achieved remarkable performance on single-turn tasks, yet their effectiveness deteriorates in multi-turn conversations. We define this phenomenon as cumulative contextual decay - a progressive degradation of contextual integrity caused by attention pollution, dilution, and drift. To address this challenge, we propose Rhea (Role-aware Heuristic Episodic Attention), a novel framework that decouples conversation history into two functionally independent memory modules: (1) an Instructional Memory (IM) that persistently stores high-fidelity global constraints via a structural priority mechanism, and (2) an Episodic Memory (EM) that dynamically manages user-model interactions via asymmetric noise control and heuristic context retrieval. During inference, Rhea constructs a high signal-to-noise context by applying its priority attention: selectively integrating relevant episodic information while always prioritizing global instructions. To validate this approach, experiments on multiple multi-turn conversation benchmarks - including MT-Eval and Long-MT-Bench+ - show that Rhea mitigates performance decay and improves overall accuracy by 1.04 points on a 10-point scale (a 16% relative gain over strong baselines). Moreover, Rhea maintains near-perfect instruction fidelity (IAR > 8.1) across long-horizon interactions. These results demonstrate that Rhea provides a principled and effective framework for building more precise, instruction-consistent conversational LLMs.

cs.CL↗

arXiv2025

Adversarial Inverse Reinforcement Learning for Mean Field Games

Mean field games (MFGs) provide a mathematically tractable framework for modelling large-scale multi-agent systems by leveraging mean field theory to simplify interactions among agents. It enables applying inverse reinforcement learning (IRL) to predict behaviours of large populations by recovering reward signals from demonstrated behaviours. However, existing IRL methods for MFGs are powerless to reason about uncertainties in demonstrated behaviours of individual agents. This paper proposes a novel framework, Mean-Field Adversarial IRL (MF-AIRL), which is capable of tackling uncertainties in demonstrations. We build MF-AIRL upon maximum entropy IRL and a new equilibrium concept. We evaluate our approach on simulated tasks with imperfect demonstrations. Experimental results demonstrate the superiority of MF-AIRL over existing methods in reward recovery.

cs.LG↗

arXiv2025

Structured Context Learning for Generic Event Boundary Detection

Generic Event Boundary Detection (GEBD) aims to identify moments in videos that humans perceive as event boundaries. This paper proposes a novel method for addressing this task, called Structured Context Learning, which introduces the Structured Partition of Sequence (SPoS) to provide a structured context for learning temporal information. Our approach is end-to-end trainable and flexible, not restricted to specific temporal models like GRU, LSTM, and Transformers. This flexibility enables our method to achieve a better speed-accuracy trade-off. Specifically, we apply SPoS to partition the input frame sequence and provide a structured context for the subsequent temporal model. Notably, SPoS's overall computational complexity is linear with respect to the video length. We next calculate group similarities to capture differences between frames, and a lightweight fully convolutional network is utilized to determine the event boundaries based on the grouped similarity maps. To remedy the ambiguities of boundary annotations, we adapt the Gaussian kernel to preprocess the ground-truth event boundaries. Our proposed method has been extensively evaluated on the challenging Kinetics-GEBD, TAPOS, and shot transition detection datasets, demonstrating its superiority over existing state-of-the-art methods.

cs.CV↗

arXiv2025

Thinking With Bounding Boxes: Enhancing Spatio-Temporal Video Grounding via Reinforcement Fine-Tuning

Spatio-temporal video grounding (STVG) requires localizing a target object in untrimmed videos both temporally and spatially from natural language descriptions. Despite their strong language understanding, multimodal large language models (MLLMs) underperform on STVG due to misaligned training objectives and weak fine-grained region-word alignment in standard visual encoders. To address this, we propose STVG-o1, the first framework that enables off-the-shelf MLLMs to achieve state-of-the-art STVG performance without any architectural modifications. Our method introduces a bounding-box chain-of-thought mechanism that explicitly reasons about spatio-temporal locations in an intermediate step before producing the final prediction. We further design a multi-dimensional reinforcement reward function consisting of format, consistency, temporal, spatial, and think rewards, which provides geometry-aware supervision through reinforcement fine-tuning. Evaluated on HCSTVG-v1/v2 and VidSTG, STVG-o1 sets new state-of-the-art results on HCSTVG, outperforming the best task-specific method by 7.3\% m\_tIoU on HCSTVG-v1, matching specialized models on VidSTG, and surpassing all existing MLLM-based approaches by large margins. It also demonstrates strong open-vocabulary generalization across datasets, establishing MLLMs as viable and powerful backbones for precise spatio-temporal grounding. Our code and models will be released.

cs.CV↗

arXiv2025

Logical operations with a dynamical qubit in Floquet-Bacon-Shor code

Quantum error correction (QEC) protects quantum systems against inevitable noises and control inaccuracies, providing a pathway towards fault-tolerant (FT) quantum computation. Stabilizer codes, including surface code and color code, have long been the focus of research and have seen significant experimental progress in recent years. Recently proposed time-dynamical QEC, including Floquet codes and generalized time-dynamical code implementations, opens up new opportunities for FT quantum computation. By employing a periodic schedule of low-weight parity checks, Floquet codes can generate additional dynamical logical qubits, offering enhanced error correction capabilities and potentially higher code performance. Here, we experimentally implement the Floquet-Bacon-Shor code on a superconducting quantum processor. We encode a dynamical logical qubit within a $3\times 3$ lattice of data qubits, alongside a conventional static logical qubit. We demonstrate FT encoding and measurement of the two-qubit logical states, and stabilize these states using repeated error detection. We showcase universal single-qubit logical gates on the dynamical qubit. Furthermore, by implementing a logical CNOT gate, we entangle the dynamical and static logical qubits, generating an error-detected logical Bell state with a fidelity of 75.9\%. Our results highlight the potential of Floquet codes for resource-efficient FT quantum computation.

quant-ph↗

arXiv2025

Thermalizing channel states for rapid qubit heating

Although known for negatively impacting the operation of superconducting qubits, thermal baths are shown to exert qubit control in a positive way, provided they are properly engineered. We demonstrate an experimental method to engineer the transduction of microwave driving into heat flow through a leaky resonator. Given the precise conversion, a qubit receiving the heat flow obtains a quasi-thermal equilibrium with arbitrary target temperature in hundreds of nanoseconds. We show that the dynamics of the quantum transducing process is described by thermalizing channel states, generated from the double dressings of the resonator by the semi-classical driving and the qubit-resonator coupling. Their spectrum, coupling, and driving strength determine the channel rate of energy flow, along with the relaxation rates of photon leakage into the bath. The analytical prediction is shown to match well with the experimental measurements on an Xmon qubit circuit.

quant-ph↗

arXiv2025

Every Activation Boosted: Scaling General Reasoner to 1 Trillion Open Language Foundation

We introduce Ling 2.0, a series reasoning-oriented language foundation built upon the principle that every activation boosts reasoning capability. Designed to scale from tens of billions to one trillion parameters under a unified Mixture-of-Experts (MoE) paradigm, Ling 2.0 emphasizes high sparsity, cross-scale consistency, and efficiency guided by empirical scaling laws. The series includes three non-thinking (instruct) models - Ling-mini-2.0, Ling-flash-2.0, and Ling-1T - ranging from 16B to 1T total parameters and achieving up to 7-fold active-compute efficiency compared with dense counterparts. Ling 2.0 integrates coordinated innovations across model architecture, pre-training, post-training, and infrastructure: a high-sparsity MoE with MTP for efficient reasoning, reasoning-oriented data and mid-training CoT activation, reinforcement-based fine-tuning (DFT, Evo-CoT), and full-scale FP8 training with fine-grained heterogeneous pipelines. At the trillion scale, Ling-1T establishes a new Pareto frontier of reasoning accuracy versus computational efficiency, demonstrating that sparse activation, when properly aligned with reasoning objectives, enables scalable and efficient intelligence. Collectively, Ling 2.0 provides a coherent, open, and efficient foundation for advancing future reasoning and thinking models, including the Ring series built upon the same base.

cs.CL↗

arXiv2025

PlanarTrack: A high-quality and challenging benchmark for large-scale planar object tracking

Planar tracking has drawn increasing interest owing to its key roles in robotics and augmented reality. Despite recent great advancement, further development of planar tracking, particularly in the deep learning era, is largely limited compared to generic tracking due to the lack of large-scale platforms. To mitigate this, we propose PlanarTrack, a large-scale high-quality and challenging benchmark for planar tracking. Specifically, PlanarTrack consists of 1,150 sequences with over 733K frames, including 1,000 short-term and 150 new long-term videos, which enables comprehensive evaluation of short- and long-term tracking performance. All videos in PlanarTrack are recorded in unconstrained conditions from the wild, which makes PlanarTrack challenging but more realistic for real-world applications. To ensure high-quality annotations, each video frame is manually annotated by four corner points with multi-round meticulous inspection and refinement. To enhance target diversity of PlanarTrack, we only capture a unique target in one sequence, which is different from existing benchmarks. To our best knowledge, PlanarTrack is by far the largest and most diverse and challenging dataset dedicated to planar tracking. To understand performance of existing methods on PlanarTrack and to provide a comparison for future research, we evaluate 10 representative planar trackers with extensive comparison and in-depth analysis. Our evaluation reveals that, unsurprisingly, the top planar trackers heavily degrade on the challenging PlanarTrack, which indicates more efforts are required for improving planar tracking. Our data and results will be released at https://github.com/HengLan/PlanarTrack

cs.CV↗

arXiv2025

Trust Region Reward Optimization and Proximal Inverse Reward Optimization Algorithm

Inverse Reinforcement Learning (IRL) learns a reward function to explain expert demonstrations. Modern IRL methods often use the adversarial (minimax) formulation that alternates between reward and policy optimization, which often lead to unstable training. Recent non-adversarial IRL approaches improve stability by jointly learning reward and policy via energy-based formulations but lack formal guarantees. This work bridges this gap. We first present a unified view showing canonical non-adversarial methods explicitly or implicitly maximize the likelihood of expert behavior, which is equivalent to minimizing the expected return gap. This insight leads to our main contribution: Trust Region Reward Optimization (TRRO), a framework that guarantees monotonic improvement in this likelihood via a Minorization-Maximization process. We instantiate TRRO into Proximal Inverse Reward Optimization (PIRO), a practical and stable IRL algorithm. Theoretically, TRRO provides the IRL counterpart to the stability guarantees of Trust Region Policy Optimization (TRPO) in forward RL. Empirically, PIRO matches or surpasses state-of-the-art baselines in reward recovery, policy imitation with high sample efficiency on MuJoCo and Gym-Robotics benchmarks and a real-world animal behavior modeling task.

cs.LG↗

arXiv2025

AwareCompiler: Agentic Context-Aware Compiler Optimization via a Synergistic Knowledge-Data Driven Framework

Compiler optimization is crucial for enhancing program performance by transforming the sequence of optimization passes while maintaining correctness. Despite the promising potential of large language models (LLMs)-based agent for software optimization, automating compiler optimization remains challenging due to: (1) semantic misalignment between abstract program representations and concrete optimization passes, (2) inefficient interaction mechanisms between agents and compiler environments, and (3) reward sparsity from the extensive decision-making process within large optimization spaces. This paper introduces \textbf{AwareCompiler}, an agentic framework for compiler optimization that addresses these challenges through three key innovations: structured knowledge integration and dataset construction, knowledge-driven adaptive pass generation, and data-driven hybrid training pipeline. Experimental results on standard benchmarks demonstrate that AwareCompiler significantly outperforms existing baselines in both performance and efficiency, highlighting the effectiveness of our synergistic knowledge-data-driven approach. Our code is publicly available at https://github.com/LHY-24/AwareCompiler.

cs.PL↗

arXiv2025

Dephasing-assisted diffusive dynamics in superconducting quantum circuits

Random fluctuations caused by environmental noise can lead to decoherence in quantum systems. Exploring and controlling such dissipative processes is both fundamentally intriguing and essential for harnessing quantum systems to gain practical advantages and deeper insights. In this work, we first demonstrate the diffusive dynamics assisted by controlled dephasing noise in superconducting quantum circuits, contrasting with coherent evolution. We show that dephasing can give distinct dynamical behavior in a superconducting qubit array with quasiperiodic order. Furthermore, by preparing different excitation distributions in the qubit array, we observe that a more localized initial state relaxes to a uniformly distributed mixed state faster with dephasing noise, illustrating another counterintuitive phenomenon called Mpemba-effect-like quantum dynamics, i.e., a far-from-equilibrium state can relax toward the equilibrium faster. These results deepen our understanding of diffusive dynamics at the microscopic level, and demonstrate controlled dissipative processes as a valuable tool for investigating Markovian open quantum systems.

quant-ph↗