arXiv subjects

Ang Li

Publications and source records attributed to Ang Li.

At least 127 records · Page 7Linked to original sources

arXiv2026

Decoupling Return-to-Go for Efficient Decision Transformer

The Decision Transformer (DT) has established a powerful sequence modeling approach to offline reinforcement learning. It conditions its action predictions on Return-to-Go (RTG), using it both to distinguish trajectory quality during training and to guide action generation at inference. In this work, we identify a critical redundancy in this design: feeding the entire sequence of RTGs into the Transformer is theoretically unnecessary, as only the most recent RTG affects action prediction. We show that this redundancy can impair DT's performance through experiments. To resolve this, we propose the Decoupled DT (DDT). DDT simplifies the architecture by processing only observation and action sequences through the Transformer, using the latest RTG to guide the action prediction. This streamlined approach not only improves performance but also reduces computational cost. Our experiments show that DDT significantly outperforms DT and establishes competitive performance against state-of-the-art DT variants across multiple offline RL tasks.

cs.AI↗

arXiv2026

Calibration-Conditioned FiLM Decoders for Low-Latency Decoding of Quantum Error Correction Evaluated on IBM Repetition-Code Experiments

Real-time decoding of quantum error correction (QEC) is essential for enabling fault-tolerant quantum computation. A practical decoder must operate with high accuracy at low latency, while remaining robust to spatial and temporal variations in hardware noise. We introduce a hardware-conditioned neural decoder framework designed to exploit the natural separation of timescales in superconducting processors, where calibration drifts occur over hours while error correction requires microsecond-scale responses. By processing calibration data through a graph-based encoder and conditioning a lightweight convolutional backbone via feature-wise linear modulation (FiLM), we decouple the heavy processing of device statistics from the low-latency syndrome decoding. We evaluate this approach using the 1D repetition code as a testbed on IBM Fez, Kingston, and Pittsburgh processors, collecting over 2.7 million experimental shots spanning distances up to d = 11. We demonstrate that a single trained model generalizes to unseen qubit chains and new calibration data acquired days later without retraining. On these unseen experiments, the FiLM-conditioned decoder achieves up to an 11.1x reduction in logical error rate relative to modified minimum-weight perfect matching. We observe that by employing a network architecture that exploits the highly asynchronous nature of system calibration and decoding, hardware-conditioned neural decoding demonstrates promising, adaptive performance with negligible latency overhead relative to unconditioned baselines.

quant-ph↗

arXiv2026

NWQWorkflow: The Northwest Quantum Workflow

This whitepaper presents NWQWorkflow, an end-to-end workflow for quantum application development, compilation, error correction, benchmarking, numerical simulation, control, and execution on a prototype superconducting testbed. NWQWorkflow integrates NWQStudio (programming GUI environment), NWQASM (intermediate representation), QASMTrans (compiler), NWQEC (quantum error correction), QASMBench (benchmarking and characterization), NWQSim (HPC simulation), NWQLib (algorithm library), NWQData (data sets), NWQControl (quantum control), and NWQSC (superconducting testbed). The system enables closed-loop software-hardware co-design and reflects the past eight years of quantum computing research the author has led at PNNL (2018-2026). By releasing most software components as open source or planning their open-source availability, we aim to cultivate a collaborative quantum information science (QIS) ecosystem and support the transition toward a scalable quantum supercomputing era.

quant-ph↗

arXiv2026

STEP3-VL-10B Technical Report

We present STEP3-VL-10B, a lightweight open-source foundation model designed to redefine the trade-off between compact efficiency and frontier-level multimodal intelligence. STEP3-VL-10B is realized through two strategic shifts: first, a unified, fully unfrozen pre-training strategy on 1.2T multimodal tokens that integrates a language-aligned Perception Encoder with a Qwen3-8B decoder to establish intrinsic vision-language synergy; and second, a scaled post-training pipeline featuring over 1k iterations of reinforcement learning. Crucially, we implement Parallel Coordinated Reasoning (PaCoRe) to scale test-time compute, allocating resources to scalable perceptual reasoning that explores and synthesizes diverse visual hypotheses. Consequently, despite its compact 10B footprint, STEP3-VL-10B rivals or surpasses models 10$\times$-20$\times$ larger (e.g., GLM-4.6V-106B, Qwen3-VL-235B) and top-tier proprietary flagships like Gemini 2.5 Pro and Seed-1.5-VL. Delivering best-in-class performance, it records 92.2% on MMBench and 80.11% on MMMU, while excelling in complex reasoning with 94.43% on AIME2025 and 75.95% on MathVision. We release the full model suite to provide the community with a powerful, efficient, and reproducible baseline.

cs.CV↗

arXiv2026

Bridging Superconducting and Neutral-Atom Platforms for Efficient Fault-Tolerant Quantum Architectures

The transition to the fault-tolerant era exposes the limitations of homogeneous quantum systems, where no single qubit modality simultaneously offers optimal operation speed, connectivity, and scalability. In this work, we propose a strategic approach to Heterogeneous Quantum Architectures (HQA) that synthesizes the distinct advantages of the superconducting (SC) and neutral atom (NA) platforms. We explore two architectural role assignment strategies based on hardware characteristics: (1) We offload the latency-critical Magic State Factory (MSF) to fast SC devices while performing computation on scalable NA arrays, a design we term MagicAcc, which effectively mitigates the resource-preparation bottleneck. (2) We explore a Memory-Compute Separation (MCSep) paradigm that utilizes NA arrays for high-density qLDPC memory storage and SC devices for fast surface-code processing. Our evaluation, based on a comprehensive end-to-end cost model, demonstrates that principled heterogeneity yields significant performance gains. Specifically, our designs achieve $752\times$ speedup over NA-only baselines on average and reduce the physical qubit footprint by over $10\times$ compared to SC-only systems. These results chart a clear pathway for leveraging cross-modality interconnects to optimize the space-time efficiency of future fault-tolerant quantum computers.

quant-ph↗

arXiv2026

Zeros can be Informative: Masked Binary U-Net for Image Segmentation on Tensor Cores

Real-time image segmentation is a key enabler for AR/VR, robotics, drones, and autonomous systems, where tight accuracy, latency, and energy budgets must be met on resource-constrained edge devices. While U-Net offers a favorable balance of accuracy and efficiency compared to large transformer-based models, achieving real-time performance on high-resolution input remains challenging due to compute, memory, and power limits. Extreme quantization, particularly binary networks, is appealing for its hardware-friendly operations. However, two obstacles limit practicality: (1) severe accuracy degradation, and (2) a lack of end-to-end implementations that deliver efficiency on general-purpose GPUs. We make two empirical observations that guide our design. (1) An explicit zero state is essential: training with zero masking to binary U-Net weights yields noticeable sparsity. (2) Quantization sensitivity is uniform across layers. Motivated by these findings, we introduce Masked Binary U-Net (MBU-Net), obtained through a cost-aware masking strategy that prioritizes masking where it yields the highest accuracy-per-cost, reconciling accuracy with near-binary efficiency. To realize these gains in practice, we develop a GPU execution framework that maps MBU-Net to Tensor Cores via a subtractive bit-encoding scheme, efficiently implementing masked binary weights with binary activations. This design leverages native binary Tensor Core BMMA instructions, enabling high throughput and energy savings on widely available GPUs. Across 3 segmentation benchmarks, MBU-Net attains near full-precision accuracy (3% average drop) while delivering 2.04x speedup and 3.54x energy reductions over a 16-bit floating point U-Net.

cs.CV↗

arXiv2026

Boosting Adversarial Transferability with Low-Cost Optimization via Maximin Expected Flatness

Transfer-based attacks craft adversarial examples on white-box surrogate models and directly deploy them against black-box target models, offering model-agnostic and query-free threat scenarios. While flatness-enhanced methods have recently emerged to improve transferability by enhancing the loss surface flatness of adversarial examples, their divergent flatness definitions and heuristic attack designs suffer from unexamined optimization limitations and missing theoretical foundation, thus constraining their effectiveness and efficiency. This work exposes the severely imbalanced exploitation-exploration dynamics in flatness optimization, establishing the first theoretical foundation for flatness-based transferability and proposing a principled framework to overcome these optimization pitfalls. Specifically, we systematically unify fragmented flatness definitions across existing methods, revealing their imbalanced optimization limitations in over-exploration of sensitivity peaks or over-exploitation of local plateaus. To resolve these issues, we rigorously formalize average-case flatness and transferability gaps, proving that enhancing zeroth-order average-case flatness minimizes cross-model discrepancies. Building on this theory, we design a Maximin Expected Flatness (MEF) attack that enhances zeroth-order average-case flatness while balancing flatness exploration and exploitation. Extensive evaluations across 22 models and 24 current transfer-based attacks demonstrate MEF's superiority: it surpasses the state-of-the-art PGN attack by 4% in attack success rate at half the computational cost and achieves 8% higher success rate under the same budget. When combined with input augmentation, MEF attains 15% additional gains against defense-equipped models, establishing new robustness benchmarks. Our code is available at https://github.com/SignedQiu/MEFAttack.

cs.CV↗

arXiv2026

Arctic-Text2SQL-R1: Simple Rewards, Strong Reasoning in Text-to-SQL

Translating natural language into SQL (Test2SQL) is a longstanding challenge at the intersection of natural language understanding and structured data access. While large language models (LLMs) have significantly improved fluency in SQL generation, producing correct and executable SQL--particularly for complex queries--remains a bottleneck. We present Arctic-Text2SQL-R1, a reinforcement learning (RL) framework and model family designed to generate accurate, executable SQL using a lightweight reward signal based solely on execution correctness. Our approach avoids brittle intermediate supervision and complex reward shaping, promoting stable training and alignment with the end task. Combined with carefully curated data, strong supervised initialization, and effective training practices, Arctic-Text2SQL-R1 achieves state-of-the-art execution accuracy across six diverse Test2SQL benchmarks, including the top position on the BIRD leaderboard. Notably, our 7B model outperforms prior 70B-class systems, highlighting the framework's scalability and efficiency. We further demonstrate inference-time robustness through simple extensions like value retrieval and majority voting. Extensive experiments and ablation studies offer both positive and negative insights, providing practical guidance for future Test2SQL research.

cs.CL↗

arXiv2026

Implementation of Tensor Network Simulation TN-Sim under NWQ-Sim

Large-scale tensor network simulations are crucial for developing robust complexity-theoretic bounds on classical quantum simulation, enabling circuit cutting approaches, and optimizing circuit compilation, all of which aid efficient quantum computation on limited quantum resources. Modern exascale high-performance computing platforms offer significant potential for advancing tensor network quantum circuit simulation capabilities. We implement TN-Sim, a tensor network simulator backend within the NWQ-Sim software package that utilizes the Tensor Algebra for Many-body Methods (TAMM) framework to support both distributed HPC-scale computations and local simulations with ITensor. To optimize the scale up in computation across multiple nodes we implement a task based parallelization scheme to demonstrate parallelized gate contraction for wide quantum circuits with many gates per layer. Through the integration of the TAMM framework with Matrix Product State (MPS) tensor network approaches, we deliver a simulation environment that can scale from local systems to HPC clusters. We demonstrate an MPS tensor network simulator running on the state-of-the-art Perlmutter (NVIDIA) supercomputer and discuss the potential portability of this software to HPC clusters such as Frontier (AMD) and Aurora (Intel). We also discuss future improvements including support for different tensor network topologies and enhanced computational efficiency.

quant-ph↗

arXiv2026

Time-Transformer: Integrating Local and Global Features for Better Time Series Generation (Extended Version)

Generating time series data is a promising approach to address data deficiency problems. However, it is also challenging due to the complex temporal properties of time series data, including local correlations as well as global dependencies. Most existing generative models have failed to effectively learn both the local and global properties of time series data. To address this open problem, we propose a novel time series generative model named 'Time-Transformer AAE', which consists of an adversarial autoencoder (AAE) and a newly designed architecture named 'Time-Transformer' within the decoder. The Time-Transformer first simultaneously learns local and global features in a layer-wise parallel design, combining the abilities of Temporal Convolutional Networks and Transformer in extracting local features and global dependencies respectively. Second, a bidirectional cross attention is proposed to provide complementary guidance across the two branches and achieve proper fusion between local and global features. Experimental results demonstrate that our model can outperform existing state-of-the-art models in 5 out of 6 datasets, specifically on those with data containing both global and local properties. Furthermore, we highlight our model's advantage on handling this kind of data via an artificial dataset. Finally, we show our model's ability to address a real-world problem: data augmentation to support learning with small datasets and imbalanced datasets.

cs.LG↗

arXiv2025

Policy-Conditioned Policies for Multi-Agent Task Solving

In multi-agent tasks, the central challenge lies in the dynamic adaptation of strategies. However, directly conditioning on opponents' strategies is intractable in the prevalent deep reinforcement learning paradigm due to a fundamental ``representational bottleneck'': neural policies are opaque, high-dimensional parameter vectors that are incomprehensible to other agents. In this work, we propose a paradigm shift that bridges this gap by representing policies as human-interpretable source code and utilizing Large Language Models (LLMs) as approximate interpreters. This programmatic representation allows us to operationalize the game-theoretic concept of \textit{Program Equilibrium}. We reformulate the learning problem by utilizing LLMs to perform optimization directly in the space of programmatic policies. The LLM functions as a point-wise best-response operator that iteratively synthesizes and refines the ego agent's policy code to respond to the opponent's strategy. We formalize this process as \textit{Programmatic Iterated Best Response (PIBR)}, an algorithm where the policy code is optimized by textual gradients, using structured feedback derived from game utility and runtime unit tests. We demonstrate that this approach effectively solves several standard coordination matrix games and a cooperative Level-Based Foraging environment.

cs.GT↗

arXiv2025

Scaling Point-based Differentiable Rendering for Large-scale Reconstruction

Point-based Differentiable Rendering (PBDR) enables high-fidelity 3D scene reconstruction, but scaling PBDR to high-resolution and large scenes requires efficient distributed training systems. Existing systems are tightly coupled to a specific PBDR method. And they suffer from severe communication overhead due to poor data locality. In this paper, we present Gaian, a general distributed training system for PBDR. Gaian provides a unified API expressive enough to support existing PBDR methods, while exposing rich data-access information, which Gaian leverages to optimize locality and reduce communication. We evaluated Gaian by implementing 4 PBDR algorithms. Our implementations achieve high performance and resource efficiency: across six datasets and up to 128 GPUs, it reduces communication by up to 91% and improves training throughput by 1.50x-3.71x.

cs.DC↗

arXiv2025

Making Large Language Models Efficient Dense Retrievers

Recent work has shown that directly fine-tuning large language models (LLMs) for dense retrieval yields strong performance, but their substantial parameter counts make them computationally inefficient. While prior studies have revealed significant layer redundancy in LLMs for generative tasks, it remains unclear whether similar redundancy exists when these models are adapted for retrieval tasks, which require encoding entire sequences into fixed representations rather than generating tokens iteratively. To this end, we conduct a comprehensive analysis of layer redundancy in LLM-based dense retrievers. We find that, in contrast to generative settings, MLP layers are substantially more prunable, while attention layers remain critical for semantic aggregation. Building on this insight, we propose EffiR, a framework for developing efficient retrievers that performs large-scale MLP compression through a coarse-to-fine strategy (coarse-grained depth reduction followed by fine-grained width reduction), combined with retrieval-specific fine-tuning. Across diverse BEIR datasets and LLM backbones, EffiR achieves substantial reductions in model size and inference cost while preserving the performance of full-size models.

cs.IR↗

arXiv2025

CoordAR: One-Reference 6D Pose Estimation of Novel Objects via Autoregressive Coordinate Map Generation

Object 6D pose estimation, a crucial task for robotics and augmented reality applications, becomes particularly challenging when dealing with novel objects whose 3D models are not readily available. To reduce dependency on 3D models, recent studies have explored one-reference-based pose estimation, which requires only a single reference view instead of a complete 3D model. However, existing methods that rely on real-valued coordinate regression suffer from limited global consistency due to the local nature of convolutional architectures and face challenges in symmetric or occluded scenarios owing to a lack of uncertainty modeling. We present CoordAR, a novel autoregressive framework for one-reference 6D pose estimation of unseen objects. CoordAR formulates 3D-3D correspondences between the reference and query views as a map of discrete tokens, which is obtained in an autoregressive and probabilistic manner. To enable accurate correspondence regression, CoordAR introduces 1) a novel coordinate map tokenization that enables probabilistic prediction over discretized 3D space; 2) a modality-decoupled encoding strategy that separately encodes RGB appearance and coordinate cues; and 3) an autoregressive transformer decoder conditioned on both position-aligned query features and the partially generated token sequence. With these novel mechanisms, CoordAR significantly outperforms existing methods on multiple benchmarks and demonstrates strong robustness to symmetry, occlusion, and other challenges in real-world tests.

cs.CV↗

arXiv2025

F2: Offline Reinforcement Learning for Hamiltonian Simulation via Free-Fermionic Subroutine Compilation

Compiling shallow and accurate quantum circuits for Hamiltonian simulation remains challenging due to hardware constraints and the combinatorial complexity of minimizing gate count and circuit depth. Existing optimization method pipelines rely on hand-engineered classical heuristics, which cannot learn input-dependent structure and therefore miss substantial opportunities for circuit reduction. We introduce F2, an offline reinforcement learning framework that exploits free-fermionic structure to efficiently compile Trotter-based Hamiltonian simulation circuits. F2 provides (i) a reinforcement-learning environment over classically simulatable free-fermionic subroutines, (ii) architectural and objective-level inductive biases that stabilize long-horizon value learning, and (iii) a reversible synthetic-trajectory generation mechanism that consistently yields abundant, guaranteed-successful offline data. Across benchmarks spanning lattice models, protein fragments, and crystalline materials (12-222 qubits), F2 reduces gate count by 47% and depth by 38% on average relative to strong baselines (Qiskit, Cirq/OpenFermion) while maintaining average errors of 10^(-7). These results show that aligning deep reinforcement learning with the algebraic structure of quantum dynamics enables substantial improvements in circuit synthesis, suggesting a promising direction for scalable, learning-based quantum compilation

quant-ph↗

arXiv2025

Exact and Efficient Stabilizer Simulation of Thermal-Relaxation Noise for Quantum Error Correction

Stabilizer-based simulation of quantum error-correcting codes typically relies on the Pauli-twirling approximation (PTA) to render non-Clifford noise classically tractable, but PTA can distort the behavior of physically relevant channels such as thermal relaxation. Physically accurate noise simulation is needed to train decoders and understand the noise suppression capabilities of quantum error correction codes. In this work, we develop an exact and stabilizer-compatible model of qubit thermal relaxation noise and show that the combined amplitude damping and dephasing channel admits a fully positive probability decomposition into Clifford operations and reset whenever $T_2 \leqslant T_1$. For $T_2 > T_1$, the resulting decomposition is negative, but allows a smaller sampling overhead versus independent channels. We further introduce an approximated error channel with reset that removes the negativity of the decomposition while achieving higher channel fidelity to the true thermal relaxation than PTA, and extend our construction to finite temperature relaxation. We apply the exact combined model to investigate large surface codes and bivariate bicycle codes on superconducting platforms with realistic thermal relaxation error. The differing logical performances across code states further indicate that noise-model-informed decoders will be essential for accurately capturing thermal-noise structure in future fault-tolerant architectures.

quant-ph↗

arXiv2025

Internal superfluid response and torque evolution in the giant glitch of PSR J1718-3718

We investigate the post-glitch rotational evolution of pulsars by analyzing the 2007 giant glitch of PSR J1718$-$3718 using a vortex creep model that incorporates both inward and outward nonlinear vortex motion, along with a time-varying external torque. A comprehensive fitting framework is developed, constrained by prior knowledge of moment of inertia participation from previous glitch studies. We apply a Markov Chain Monte Carlo approach to quantify uncertainties and parameter correlations. The model reproduces the observed timing data and yields physically consistent values for moment of inertia fractions and creep timescales. Our results indicate that inward creep and a long-term change in external torque dominate the observed increase in spin-down rate, pointing to structural changes within the star-likely triggered by a crustquake that initiated both vortex motion and a change in the moment of inertia. We estimate that the glitch involved approximately $2.4 \times 10^{12}$ inward-moving vortices and $\sim 142$ crustal plates with a typical size of $\sim 0.03$ km. This study demonstrates that detailed post-glitch modeling of sparse timing data can simultaneously constrain internal superfluid dynamics and external torque evolution, providing a quantitative framework to probe the structural properties of neutron star interiors.

astro-ph.HE↗

arXiv2025

Understanding and Harnessing Sparsity in Unified Multimodal Models

Large multimodal models have achieved remarkable progress in both understanding and generation. Recent efforts pursue unified multimodal models that integrate heterogeneous components to support both capabilities within a single framework. However, such unification introduces inference inefficiencies, e.g., specific tasks or samples may not require the full knowledge or capacity of the unified model. Yet, a systematic understanding of how these inefficiencies manifest across different components remains limited. In this work, we first conduct a systematic analysis of unified multimodal model components using training-free pruning as a probing methodology, considering both depth pruning and width reduction. Our study reveals that the understanding component exhibits notable compressibility in both understanding and generation tasks, which is more pronounced in the latter. In contrast, the generation components are highly sensitive to compression, with performance deteriorating sharply even under moderate compression ratios. To address this limitation, we propose the Mixture-of-Experts (MoE) Adaptation, inspired by the dynamic activation patterns observed across different samples. This approach partitions the generation module into multiple experts and enables sparse activation to restore generation quality. We validate the effectiveness of sparse activation through expert-frozen tuning and further demonstrate that a fully trainable adaptation delivers additional gains. As a result, the adapted BAGEL model achieves performance comparable to the full model while activating only about half of its parameters. The code is released at \href{https://github.com/Shwai-He/SparseUnifiedModel}{this link}.

cs.CV↗