arXiv subjects

Haibing Guan

Publications and source records attributed to Haibing Guan.

At least 19 recordsLinked to original sources

arXiv2026

Channel-Level Mask Guidance for Slimmable Image Compression

Learned image compression (LIC) has achieved promising rate-distortion performance, but most methods lack scalability for diverse platforms with varying computational and storage constraints. Existing slimmable LIC methods achieve flexible complexity via consecutive-channel activation but suffer from two limitations: forced consecutive-channel selection degrades narrow-model performance at low bitrates, and the lack of stable variable-rate control limits adaptability to varying bandwidth. We propose a channel-level mask-guided slimmable image compression method. A learnable channel-wise mask with sparsity-promoting regularization enables non-contiguous top-w channel selection, letting narrow models use the most informative channels. After training, mask merging and channel reordering convert the model into a standard slimmable network via Mask-to-Slimmable (M2S) conversion, enabling efficient inference without architectural changes. For variable-rate control, we combine lambda- and width-conditioned modulation with quantization-step scaling and a progressive training pipeline stabilized by a freeze-then-fine-tune strategy. Experiments on Kodak across E2E, Hyperprior, and Checkerboard backbones show superior rate-distortion performance across widths and bitrates, with large gains for narrow models at low bitrates. At w=48, it achieves up to 19.22%, 24.11%, and 22.76% BD-rate savings over the same-width Slimmable baseline on E2E, Hyperprior, and CKBD, respectively. A single VBR model replaces 40 independent models, reducing storage by 14.8x to 15.9x. Code: https://github.com/sjtu-tcloud/MGSIC.

eess.IV↗

arXiv2026

GUMPIM: Unitary and Malleable Memory for Processing-in-Memory with Guaranteed PIM Pages

DRAM-based Processing-in-Memory (PIM) addresses the "memory wall" by executing computations directly inside main memory. However, memory interleaving and virtual memory limit contiguous data size visible to PIM units, constraining PIM task granularity. To mitigate this, existing PIM systems drastically isolate PIM memory or disable memory interleaving. These design choices, however, decrease the CPU memory bandwidth and introduce extra data transfer, leading to an additional "system memory wall" that must be resolved to realize PIM's full potential. In this work, we propose GUMPIM, a PIM system that allows interleaved CPU pages and non-interleaved PIM pages to coexist in a Unitary and Malleable memory space with Guaranteed PIM page allocation. GUMPIM enables zero-copy during PIM task offloading and maintains CPU memory bandwidth while ensuring low PIM offloading overhead. First, we propose a dual-track memory management mechanism consisting of independent page allocation and address translation for CPU and PIM pages. Second, we design GUMPIM interface hardware on PIM-enabled DRAMs to provide a dynamic address mapping for the different data layouts of CPU and PIM pages. Third, we propose a PIM-assisted page migration mechanism that transparently migrates pages while preserving CPU access bandwidth, thereby enabling guaranteed and accelerated PIM page allocation. GUMPIM requires no changes to commodity DRAM standards; all hardware modifications are limited to the DRAM side, ensuring full compatibility with existing CPUs and enabling immediate deployment on current HBMx- and LPDDRx-based PIM platforms. Our results show only <0.1% performance degradation for CPU workloads on GUMPIM. For PIM workloads, GUMPIM reduces memory allocation and CPU-part computation times by 2.7x and 4.93x, respectively, yielding an end-to-end 2.3x speedup over a state-of-the-art baseline system.

cs.AR↗

arXiv2026

SWE-at-Scale: Scalable Issue Synthesis from Historical Bug Fixes

As coding agents become increasingly capable, robust evaluation requires a broader and more diverse supply of verifiable software engineering (SWE) tasks. However, real-world tasks are constrained by the natural frequency of issue reporting and the substantial cost of manual curation. Furthermore, existing synthetic approaches rely primarily on the pretrained knowledge of large models coupled with heuristic prompting, often failing to capture the authentic, complex mechanics of software failures. To exponentially scale SWE task production, we introduce SWE-at-Scale, a framework that automatically synthesizes realistic issues by reverse-engineering real-world software fixes. Our key insight is that historical fixes encapsulate deep mechanistic knowledge of bug formation: inverting a fix transforms repair supervision into bug-induction supervision grounded in actual development scenarios. SWE-at-Scale transfers these learned bug patterns to novel code contexts, repeatedly sampling diverse bug typologies to construct executable SWE tasks. Extensive experiments show that SWE-at-Scale generates more diverse and realistic SWE tasks than current synthesis baselines while posing greater challenges to state-of-the-art coding agents. Notably, SWE-at-Scale can generate distinct bug types from a single code context, with bug diversity scaling robustly even up to 200 generations per file. Overall, SWE-at-Scale turns finite real-world fixes into reusable supervision for scalable synthesis of realistic, diverse, and challenging SWE tasks.

cs.SE↗

arXiv2026

One Analyst Is Not Ground Truth: Grading Agent-Built Financial Models Against Observed Professional Practice

Language model agents can now construct complete financial models, but it remains unclear whether they have learned the financial judgment that gives those models meaning. Existing benchmarks often anchor correctness to an expert-authored solution (e.g., numerical targets or detailed rubrics) which introduces a implicit assumption: \emph{one expert solution can serve as ground truth.} This is appropriate when finance provides a unique answer, but not when judgment is required. Evidence from professional practice challenges this assumption: When financial models built by different analysts for the same company are graded against one another, the median score is only 0.33 under standard tolerances, \emph{revealing that single-reference grading confounds professional disagreement with error.} We therefore introduce GAUGE, a benchmark that decomposes financial-model evaluation into deterministic checks, rubric judgments and numerical rules. Built from 1,001 professional valuation spanning 922 companies and all 25 GICS industry groups, GAUGE checks mechanical properties deterministically and evaluates judgment-bearing quantities against ranges in professional practice. We then validate GAUGE as a measurement instrument by testing expertise ordering, held-out professional values, and robustness to LM judgements. Our findings show that current LM agents are far better at constructing financial models than at deriving the company-specific assumptions that drive valuation, a gap that even persists after fine-tuning.

cs.LG↗

arXiv2026

Certainty Is Not Just Correctness: Rethinking Token-Level Certainty in LLM Reasoning

Token-level certainty is widely used as a proxy for correctness in LLM training and inference. However, the performance of certainty-based methods depends both on the information in certainty scores and on how those scores are used. We therefore directly assess certainty's predictive ability through controlled empirical evaluations across models and tasks. We distinguish two prediction targets: identifying questions a model is more likely to answer correctly and distinguishing correct from incorrect responses to the same question. In our experiments, certainty is generally better at identifying questions a model is likely to answer correctly than at distinguishing correct from incorrect responses to the same question. Certainty also varies systematically across token types and positions within words, reflecting local properties of words and text form. Information about question difficulty appears early in generation, while the weaker information about answer correctness is more concentrated near the end. These findings show that the information certainty provides for decisions depends on the prediction target, the model, the certainty metric, and which token positions in the response are included in aggregation. We further demonstrate the practical value of these findings for test-time compute. We allocate the number of responses using certainty early in generation and weight answer votes using certainty near the end of each response. Compared with a fixed-sampling majority-voting baseline, this approach increases overall accuracy from 78.71\% to 79.54\% while reducing generated-token cost by 82.4\%.

cs.CL↗

arXiv2026

FSAN: Flow State Attention Network for Aerodynamic Prediction

Accurate aerodynamic prediction is critical for designing fuel-efficient and safe transportation systems such as aircraft and automobiles, yet traditional computational fluid dynamics (CFD) simulations remain computationally expensive and expertise-intensive, severely limiting their use in iterative design and real-time analysis. Existing deep learning surrogates suffer from two major limitations: (i) they are evaluated on datasets with narrow flow-condition ranges, leaving their performance under complex flow conditions undemonstrated; (ii) they treat global flow conditions as a single vector injected uniformly across all surface points, ignoring that different geometric regions experience distinct local flow phenomena, which degrades prediction accuracy under complex flow conditions. To address these limitations, we propose the Flow State Attention Network (FSAN). FSAN separately encodes point cloud and flow conditions, then partitions the geometry into multiple flow states via learnable soft assignments, and uses flow features to update these state representations, which in turn influence point cloud features through state changes. This enables fine-grained, state-specific interaction between geometry and flow information. Extensive experiments on two well-recognized aerodynamic benchmarks demonstrate that FSAN achieves the highest accuracy among the methods compared in this work at a higher computational cost. On Emmi-Wing, FSAN reduces the Relative L2 (REL-L2) error by over 20\% compared to the strongest baseline (Transolver), and on DrivAerNet++, it achieves a 10\% reduction compared to the strongest baseline (AdaField). These results establish FSAN as a promising neural surrogate on public benchmarks with diverse flow conditions and geometries.

cs.CV↗

arXiv2026

InSituMeasure: Probing Situated Measurement Grounding in Industrial Scenes with Multimodal Large Language Models

For trained operators, gauge reading requires little specialized knowledge, low cognitive effort, and high repeatability. Yet Multimodal Large Language Models (MLLMs) remain unreliable in continuous-valued measurement despite strong results on general multimodal benchmarks. Existing benchmarks expose this weakness but isolate measurement from realistic, knowledge-grounded settings, with limited situated context, specialized instruments, real-world noise, and matched diagnostic annotations, reducing realism and constraining root-cause analysis. We introduce InSituMeasure to evaluate situated measurement grounding. It contains 2,922 real industrial monitoring scenes across eight functional categories of professional engineering instruments, with dense gauge-attribute annotations and noise tags for failure diagnosis. We define metrics for numerical accuracy under predefined tolerances and unit consistency, rejection of fake or unanswerable tasks, and alignment between model failures and annotated error factors. Across 24 state-of-the-art MLLMs, the best model reaches only 25.7\% joint value-unit accuracy and 51.8\% confidence-diagnosis F1, revealing a substantial gap between general multimodal competence and reliable situated measurement. Further analysis identifies failures from text-induced shortcuts, overconfident responses, and authentic industrial noise, including mixed disturbances, viewpoint deviation, occlusion, and environmental interference.

cs.AI↗

arXiv2026

PCoMoE: Shifting MoE Inference from Monolithic Expert Selection to Fine-Grained Path Composition

Mixture-of-Experts (MoE) architectures scale Large Language Model (LLM) capacity efficiently by activating a sparse subset of experts per token. However, modern MoE inference remains heavily constrained by the rigid, whole-expert abstraction. Existing frameworks manage, schedule, or prune experts as atomic execution units, which fixes the optimization boundary too early and leaves fine-grained intra-expert computational redundancy underexplored. In this work, we present PCoMoE, a path-compositional execution framework that shifts MoE inference from coarse-grained expert selection to fine-grained path composition. PCoMoE incorporates a path-level formulation of expert computation, a compatibility-aware layer-wise pruning strategy to suppress low-value path combinations, and a hardware-friendly execution engine to exploit reusable sub-expert structures under strictly bounded overheads. Experimental results demonstrate that PCoMoE achieves up to a 1.31x end-to-end inference speedup while enhancing model accuracy by 10%. The code is available at https://github.com/gzyyy0/PCoMoE

cs.CL↗

arXiv2026

To Think or Not to Think: Allocating Reasoning Where It Helps

Reinforcement learning (RL) has proven effective in enhancing the reasoning performance of large language models (LLMs), particularly in complex mathematical and programming tasks. However, this capability comes with systematic \textit{length misallocation}, in which models devote excessive reasoning to simple questions while terminating prematurely on harder ones, degrading inference efficiency with negligible accuracy improvement. Many length-adaptive methods mitigate this issue by allocating token budgets according to question difficulty, under the implicit assumption that harder questions benefit monotonically from extended reasoning. In contrast, we find that the effect of reasoning length on accuracy is concentrated on \textit{partially solvable} questions. Our further analysis reveals that explicit length rewards can produce unintended training dynamics. Motivated by these findings, we propose \textbf{CARE}---\textbf{C}ontrastive \textbf{A}ccuracy \textbf{R}eward \textbf{E}stimation---which compares the beneficial length adjustment per question from online sampled responses and applies adaptive length rewards within Group Relative Policy Optimization, with no extra hyperparameters or additional inference cost. Experiments across multiple reasoning benchmarks demonstrate that our method improves Pass@1 by up to $4\%$ while simultaneously reducing reasoning length by $37\%$, achieving higher token efficiency. Code will be available upon the acceptance of this paper.

cs.AI↗

arXiv2026

FPGAgent: An LLM-Assisted Framework for Autonomous HLS Code Generation and Verification in FPGA Environments

Large language models (LLMs) have shown substantial promise for high-level synthesis (HLS) code generation, but most existing approaches validate only simulation or synthesis results. Because of timing and place-and-route constraints, \emph{HLS code that passes simulation and synthesis may still fail to produce deployable, runnable designs on real FPGA platforms}. Moreover, the lack of public benchmarks has limited many evaluations to small, self-curated test suites. We propose FPGAgent, a multi-agent framework tailored to real FPGA environments for autonomous HLS coding with end-to-end executability validation. To the best of our knowledge, FPGAgent is \emph{the first task-specification-to-executable HLS generation framework experimentally validated on a well-established benchmark}. Given a natural-language task specification, FPGAgent injects HLS-specific knowledge and employs evolutionary search to iteratively derive reliable HLS kernel implementations. It then generates a C++ validation program to verify functional correctness, diagnoses potential defects, and guides targeted repairs. Finally, it synthesizes host code for compilation and board-level execution on FPGA hardware. We comprehensively evaluate FPGAgent with five established LLMs on HLS-Eval, a benchmark containing 78 tasks across multiple domains, and verify board-level executability on a real FPGA platform. Compared with existing baselines, FPGAgent improves the synthesizable rate by 16.9% on average, executability by 26.7%, and functional correctness by 30.6%. These results show that FPGAgent substantially improves the practical usability of LLM-based HLS generation and demonstrates the value of end-to-end validation.

cs.SE↗

arXiv2026

HIERA: Workload-Aware Planning Across Implementation Spaces for GPU Kernel Optimization

High-performance GPU kernels underpin modern deep learning and scientific computing. As workloads become increasingly diverse and GPU hardware evolves rapidly, developing efficient methods for automated GPU kernel generation and optimization has become increasingly important. Existing LLM-based methods typically optimize within a fixed implementation space, limiting either optimization flexibility or search efficiency. We propose \textsc{HIERA}, a hierarchical search-space planning framework for GPU kernel optimization. \textsc{HIERA} constructs contract-augmented task specifications, selects an appropriate implementation space across PyTorch operators, CUDA libraries, and custom CUDA kernels, and uses profiling feedback and expert knowledge to guide structured iterative refinement. Experiments on KernelBench across multiple various workload levels and base LLMs show that \textsc{HIERA} delivers stronger overall implementation validity, sample efficiency, and optimization performance than existing training-free methods, while remaining competitive with the training-based CUDA-L1 without additional model training. A case study on a specialized stencil operator from scientific computing further achieves a $1.53\times$ speedup over cuDNN, demonstrating the potentiality of the general framework beyond standard machine-learning workloads.

cs.DC↗

arXiv2026

Schrödinger's Code Repository: Have LLMs Learned SWE-bench or Memorized It?

Repository-level coding benchmarks have become the standard for evaluating coding agents, yet they inherently suffer from data leakage because they are built upon popular open-source repositories repeatedly used for training. Consequently, strong performance may reflect memorization of canonical repository cues rather than robust repository reasoning. We propose SchrodingerRepo (Schrödinger's Repository), an evaluation framework for testing coding agents under dynamically instantiated repository representations. Instead of repeatedly using a static representation of the test repository, SchrodingerRepo treats the test repository as an evaluation-time latent variable that is dynamically instantiated only when the agent enters the evaluation environment. The instantiated repository preserves the original executable behavior while eroding familiar cues such as naming conventions, file layouts, and implementation patterns through four transformation levels: problem statement reconstruction, namespace remapping, intra-file layout reordering, and functionality-preserving code rewriting. We evaluate popular LLMs on SWE-bench Verified and SWE-QA. Results show that removing familiar repository cues consistently degrades agent performance and substantially increases interaction costs across models. Further analysis reveals that the additional cost is primarily caused by increased difficulty in repository exploration and localization. These findings suggest that current coding agents may partially rely on memorized repository-side cues, highlighting the need for evaluation under dynamically instantiated repository representations.

cs.SE↗

arXiv2026

Repo0: Design-Driven Zero-to-All Code Generation

Large language model agents have made substantial progress in code generation, yet most existing systems assume a predefined repository architecture. This assumption does not hold in zero-to-all code generation, where an agent must construct an entire software project directly from natural-language requirements while maintaining a modular repository architecture throughout development. We present Repo0, a continuous structural evolution framework for zero-to-all code generation. Repo0 maintains an explicit architectural state instantiated as a Dual-Directed-Acyclic-Graph (Dual-DAG), consisting of a requirement-level DAG, a component-level DAG, and their alignment relation. Starting from natural-language requirements, it iteratively evolves component boundaries through structural actions guided by modularity metrics until structural convergence, after which the converged architecture guides test-driven development code generation. We evaluate Repo0 on six real-world repositories from RepoCraft using GPT-5 mini and DeepSeek V3.2. Repo0 achieves the highest Functionality Coverage and Pass Rate across all settings. Compared with RPG, the strongest repository-planning baseline, Repo0 improves Functionality Coverage by up to 20.08 percentage points and Pass Rate by up to 29.74 percentage points. Ablation and structural-evolution analyses further demonstrate the importance of the Dual-DAG architectural state, modularity-guided structural evolution, and explicit structural convergence.

cs.SE↗

arXiv2026

SkillForge: Self-Distilling Agents for Project-Specific Issue Resolution

Large language model (LLM) based agents have demonstrated remarkable proficiency in automated software issue resolution, yet they often struggle to resolve issues in a specific repository because they lack project-specific knowledge. Existing self-evolving approaches acquire such knowledge from repository history or online repair trajectories, but they either depend on available historical issue-resolution signals or incur substantial per-issue test-time exploration cost. In this paper, we propose SkillForge, a self-distillation framework that proactively acquires project-specific knowledge from the repository itself. Instead of waiting for real issues to expose project-specific knowledge gaps, SkillForge synthesizes project-specific issues by re-implementing test-covered core functionalities of the repository. By resolving these synthetic issues, SkillForge distills reusable project-specific knowledge into entity-grounded skills and associates them with relevant repository entities for future issue resolution. Extensive experiments using both open-source and closed-source models show that SkillForge consistently improves issue resolution performance over strong baselines. These results demonstrate that proactively acquiring project-specific knowledge before solving real issues substantially improves downstream software issue resolution.

cs.SE↗

arXiv2026

M3TR: Temporal Retrieval Enhanced Multi-Modal Micro-video Popularity Prediction

Accurately predicting the popularity of micro-videos is a critical but challenging task, characterized by volatile, `rollercoaster-like' engagement dynamics. Existing methods often fail to capture these complex temporal patterns, leading to inaccurate long-term forecasts. This failure stems from two fundamental limitations: \ding{172} a superficial understanding of user feedback dynamics, which overlooks the mutually exciting and decaying nature of interactions such as likes, comments, and shares; and~\ding{173} retrieval mechanisms that rely solely on static content similarity, ignoring the crucial patterns of how a video's popularity evolves over time. To address these limitations, we propose \textbf{M$^3$TR}, a \textbf{T}emporal \textbf{R}etrieval enhanced \textbf{M}ulti-\textbf{M}odal framework that uniquely synergizes fine-grained temporal modeling with a novel temporal-aware retrieval process for \textbf{M}icro-video popularity prediction. At its core, M$^3$TR introduces a Mamba-Hawkes Process (MHP) module to explicitly model user feedback as a sequence of self-exciting events, capturing the intricate, long-range dependencies within user interactions (for \textbf{limitation} \ding{172}). This rich temporal representation then powers a temporal-aware retrieval engine that identifies historically relevant videos based on a combined similarity of both their multi-modal content (visual, audio, text) and their popularity trajectories (for \textbf{limitation} \ding{173}). By augmenting the target video's features with this retrieved knowledge, M$^3$TR achieves a comprehensive understanding of prediction. Extensive experiments on two real-world datasets demonstrate the superiority of our framework. M$^3$TR achieves state-of-the-art performance, outperforming previous methods by up to \textbf{19.3}\% in nMSE and showing significant gains in addressing long-term prediction challenges.

cs.MM↗

arXiv2026

Algorithm-Architecture Co-Design for Efficient VLA Inference via Speculative Inference and Verification

Vision-Language-Action (VLA) models have demonstrated remarkable capabilities in the field of embodied AI, but their high computational cost and limited predicted action length hinder real-time deployment. Although Dadu-Corki, a dedicated accelerator for efficient embodied AI, has been introduced, it does not exploit the inherent interaction patterns between the robot and its environment, which results in a relatively short predicted action length. We observe that robotic environments naturally alternate between active states-where precise actions are crucial-and inactive states-where actions have limited impact on task success. This insight enables a new scheduling opportunity: long-action-length speculative prediction in inactive states, paired with selective verification in active states. We propose SpecVLA, an algorithm-system co-design framework that adaptively balances action length, inference latency, and task reliability. On the algorithm side, SpecVLA introduces a state-aware VLA inference execution paradigm and a hardware-friendly construction of a smaller verification model (sVLA) using differential residuals and block-wise mixed-precision quantization. On the system side, we develop a heterogeneous architecture consisting of a GPU and a robotic-specific hardware module, along with a speculative dataflow that decouples VLA and sVLA through parallel execution. Comprehensive evaluations on OpenVLA and RDT across LIBERO and ManiSkill benchmarks show that SpecVLA reduces end-to-end latency significantly while preserving task success rate. By enabling long-action-length speculative prediction with timely verification, SpecVLA achieves real-time robotic manipulation with both high efficiency and reliability.

cs.RO↗

arXiv2026

A Motion-Aware Vector Quantization Framework with Centroid Reuse for Efficient VLA Inference

Vision-Language-Action (VLA) models have demonstrated strong potential for embodied AI, yet their high inference latency on GPUs limits real-time deployment. Existing accelerators, such as Dadu-Corki, improve efficiency but treat VLA models as full-precision workloads, leaving substantial redundancy in both memory and computation underexploited. In this paper, we propose VQVLA, an algorithm-hardware co-design framework that accelerates VLA inference by exploiting weight similarity and execution dynamics. We first introduce MotionVQ, a motion-aware vector quantization scheme that dynamically adjusts quantization precision based on the robot's execution state, reducing memory access while preserving task success rate. We then propose a merged-centroid vectorized GEMM paradigm that operates on the codebook-index representation, eliminating redundant multiplications through spatial aggregation and temporal reuse of centroids. To realize these optimizations, we design an accelerator that efficiently supports dynamic precision selection and centroid-reuse computation. Experimental results show that VQVLA achieves 6.5x, 2.8x, 1.9x, 3.3x, and 4.3x speedup over the A100 GPU, Dadu-Corki, LUT-DLA, CodeGEMM, and ShiftAddLLM, respectively, with negligible accuracy degradation.

cs.AI↗

arXiv2026

FedCod: An Efficient Communication Protocol for Cross-Silo Federated Learning with Coding

Federated Learning (FL) is an innovative distributed machine learning paradigm that enables multiple parties to collaboratively train a model without sharing their raw data, thereby preserving data privacy. Communication efficiency concerns arise in cross-silo FL, particularly due to the network heterogeneity and fluctuations associated with geo-distributed silos. Most existing solutions to these problems focus on algorithmic improvements that alter the FL algorithm but sacrificing the training performance. How to address these problems from a network perspective that is decoupled from the FL algorithm remains an open challenge. In this paper, we propose FedCod, a new application layer communication protocol designed for cross-silo FL. FedCod transparently utilizes a coding mechanism to enhance the efficient use of idle bandwidth through client-to-client communication, and dynamically adjusts coding redundancy to mitigate network bottlenecks and fluctuations, thereby improving the communication efficiency and accelerating the training process. In our real-world experiments, FedCod demonstrates a significant reduction in average communication time by up to 62% compared to the baseline, while maintaining FL training performance and optimizing inter-client communication traffic.

cs.DC↗