Searcharxiv⌕ Search

arXiv subjects

Feng Yu

Publications and source records attributed to Feng Yu.

At least 19 recordsLinked to original sources
arXiv2026

Joint optimisation of interim timing and stopping boundaries in group sequential trials

Group sequential designs (GSDs) are common in confirmatory trials, and their efficiency depends jointly on the timing of interim analyses and the shape of stopping boundaries. Most existing approaches optimise one component while holding the other fixed. Joint optimisation over free critical values has been studied for two- and three-stage efficacy-only designs and for tests with binding futility boundaries. Here we jointly optimise the full interim schedule and the boundary shape within the Hwang-Shih-DeCani and Kim-DeMets spending families, minimising the expected sample size under a specified alternative while controlling overall type I and type II error rates. A convex error-spending constraint prevents excessive early type I error spending. On the canonical information scale, the joint optimum depends only on the error rates, number of analyses and spending family, and is invariant to endpoint type and effect size. Without the convexity constraint, the parametric optima reproduced the unrestricted optima of Brittain & Bailey (1993) to within 0.001 in expected inflation factor. For three or more looks the jointly optimal design Pareto-dominated the scheduling-only optimal Pocock design, achieving the same or smaller expected sample size with uniformly smaller maximum sample size. Under a prespecified 1% tolerance, six analyses, five interim plus a final, captured essentially all efficiency attainable with up to seven analyses, reducing expected sample size by 23% and 33% at 80% and 90% power relative to the fixed-sample design. We illustrate the approach by redesigning the CORRECT trial.

stat.ME↗
arXiv2026

MatOS: An Agent Operating System for Materials Research

Materials research relies on scattered and growing data, knowledge, and simulation tools, and a single study often spans multi-day workflows across literature analysis, computation, and experiment. Large language model agents can now retrieve materials data, operate simulation codes, and close design-synthesis-characterization loops with autonomous laboratories. However, extending these agents with new capabilities, keeping long-running tasks on track, and letting them learn from past tasks and user feedback remain difficult. Here we present MatOS, an agent operating system for materials research with three core components: (1) a layered scientific asset graph that organizes data, knowledge, tools, skills, roles, and workflows as reusable assets; (2) a stateful adaptive workflow that runs tasks as explicit execution graphs with scientific checks, so that they can be resumed and traced; and (3) evidence-driven asset evolution, which revises these assets from execution records and user feedback. On ResearchClawBench, MatOS improves on the previously best-performing agent by 11.15% with the same base model and extends the cost-score Pareto frontier. With a common base model, it outperforms Codex CLI and Claude Code, exceeding Codex CLI by 2.13% on MatTools and 4.17% on SciAgentGym. In eight research cases, MatOS narrowed more than 30 million candidate structures to 97 solid-electrolyte candidates, identified 89 halide variants with higher computed conductivity than their parent structures, linked the force errors of 11 pretrained interatomic potentials to local atomic environments, and supported two new methods that reduce errors by 3.1% and 13.6% relative to their baselines. These results show that organizing research capabilities as reusable, evolving assets enables agents to carry out long, multistage materials studies. Code is available at https://github.com/DingyangLyu/MatOS.

cs.AI↗
arXiv2026

Understanding compression-extension error asymmetry in universal machine-learning interatomic potentials

Universal machine-learning interatomic potentials are increasingly used across diverse materials, yet their reliability under directional deformation remains unclear. We compare compression and extension along paired deformation paths from the same parent structures for eleven pretrained potentials. The benchmark comprises 1,963 Materials Project trajectory (MPtrj) parents and 1,286 experimental-source MinX-Eq parents at six strain magnitudes from 1% to 10%. Density-functional theory provides references for isotropic, uniaxial and biaxial loading in MPtrj and isotropic loading in MinX-Eq. At $\pm 10\%$, every model has a larger median force error under compression, with pooled isotropic compression-to-extension ratios of 2.12 and 3.30 for MPtrj and MinX-Eq. Decomposing error growth into DFT force-scale growth and relative-error evolution shows that relative errors decrease on both branches, faster under compression in MPtrj but slower under compression in MinX-Eq. Force-scale matching moves the error ratios toward unity, although MinX-Eq retains a compression-side excess across models. Radial and SOAP analyses place compressed structures at shorter separations and farther from public reference pools, while readout-associated model features show larger compression-side displacements. These results distinguish stronger physical responses from changes in relative predictive accuracy and provide a basis for assessing universal potentials in deformation simulations.

cond-mat.mtrl-sci↗
arXiv2026

RPP: A Certified Poisoned-Sample Detection Framework for Backdoor Attacks under Dataset Imbalance

Deep neural networks are highly susceptible to backdoor attacks, yet most defense methods to date rely on balanced data, overlooking the pervasive class imbalance in real-world scenarios that can amplify backdoor threats. This paper presents the first in-depth investigation of how the dataset imbalance amplifies backdoor vulnerability, showing that (i) the imbalance induces a majority-class bias that increases susceptibility and (ii) conventional defenses degrade significantly as the imbalance grows. To address this, we propose Randomized Probability Perturbation (RPP), a certified poisoned-sample detection framework that operates in a black-box setting using only model output probabilities. For any inspected sample, RPP determines whether the input has been backdoor-manipulated, while offering provable within-domain detectability guarantees and a probabilistic upper bound on the false positive rate. Extensive experiments on five benchmarks (MNIST, SVHN, CIFAR-10, TinyImageNet and ImageNet10) covering 10 backdoor attacks and 12 baseline defenses show that RPP achieves significantly higher detection accuracy than state-of-the-art defenses, particularly under dataset imbalance. RPP establishes a theoretical and practical foundation for defending against backdoor attacks in real-world environments with imbalanced data.

cs.CR↗
arXiv2026

Origins of Universal Machine Learning Force-Field Errors in Multicomponent Materials

Universal machine learning force-field generalization to multicomponent environments generated by compositional design remains insufficiently assessed. We construct a benchmark of 7,599 multicomponent configurations inspired by high-entropy design, elemental substitution and anion mixing. Eleven pretrained models are evaluated against density functional theory for energies, forces and stresses, with assessment extended to elastic, vibrational and adsorption-related properties. Force errors are analysed through training-reference coverage, local geometric heterogeneity, distance directionality and elemental response. Distances to training-reference environments reveal a qualitative association between coverage differences and increasing errors, while substantial variation remains at similar distances. Higher-error groups show greater local geometric heterogeneity, although OMat24 provides broad coverage of these environments. Relative to training-reference pair medians, errors remain low near the median, rise steeply on the compression side and increase more weakly on the extension side. After matching element pairs and absolute distance deviations, compression-side force errors are 1.81-1.95 times extension-side errors. Model-predicted pairwise interaction curves show greater curvature under compression. Fitting difficulty in independent elemental systems correlates with electronic band-energy responses to atomic displacements and Fermi-level shifts, and a similar pattern is observed in multicomponent systems. In parameter-matched comparisons, spherical-harmonic representations with maximum degrees of 2 and 4 lower test force errors for 38 and 40 of 43 elements, respectively, while differences in elemental difficulty remain. These findings inform force-field selection for experimental compositional design and identify targets for training-data sampling and model representations.

cond-mat.mtrl-sci↗
arXiv2026

Efficient 3D Whole-Body PET Image Denoising via Conditional Rectified Flow With Optimized Sampling Strategy

Reducing radiation exposure in Positron Emission Tomography (PET) is important for patient safety; however, ultra-low-dose imaging suffers from severe noise, which may affect diagnostic interpretation without appropriate image enhancement. While current 3D deep generative models, particularly diffusion models, have shown strong reconstruction fidelity, their practical use can be limited by long inference times. In contrast, faster 2D-based alternatives may have difficulty maintaining volumetric consistency, an important consideration for whole-body PET imaging analysis. To bridge this gap, we propose a one-pass conditional 3D rectified flow (3D Flow) framework for whole-body PET image denoising that incorporates a novel optimized non-uniform sampling strategy. The model is trained with a one-pass linear-interpolant velocity-matching objective. This approach reconstructs a full 3D volume in approximately 30 seconds in our implementation, compared with multi-hour inference for the evaluated 3D DDPM baseline. Evaluations including zero-shot transfer to an independent clinical dataset show that our model achieves favorable global image quality and lesion conspicuity compared with the evaluated 3D DDPM and DDIM baselines, including on challenging short-acquisition data. Furthermore, the proposed method shows promising zero-shot transfer performance across the evaluated datasets and unseen dose levels (down to 1/100 of the standard dose), with artifact-focused visual comparisons supporting the need for further lesion-level validation. By balancing reconstruction fidelity and computational efficiency, this work presents a candidate approach for ultra-low-dose whole-body PET image denoising.

cs.CV↗
arXiv2026

HLS-Seek: QoR-Aware Code Generation for High-Level Synthesis via Proxy Comparative Reward Reinforcement Learning

High-Level Synthesis (HLS) compiles algorithmic C/C++ descriptions into hardware, with Quality of Results (QoR)---latency and resource utilization---critically governed by pragma configurations and code structure. Existing natural-language-to-HLS (NL-to-HLS) training approaches prioritize functional correctness while largely ignoring QoR. We observe that reinforcement learning (RL) for HLS does not require absolute synthesis results---only relative comparisons between candidates. Based on this insight, we propose \textbf{HLS-Seek}, a QoR-aware NL-to-HLS framework that avoids full synthesis-in-the-loop RL via a comparative proxy reward model achieving 99.53\% Pareto-dominance accuracy. To prevent reward hacking, we introduce \textit{uncertainty-aware Monte Carlo (MC) dropout switching} that selectively invokes real Vitis HLS synthesis for low-confidence candidates and online updates the proxy, creating a self-improving reward system. HLS-Seek achieves 84.7\% syntax correctness pass@1 and 81.4\% functional correctness pass@5 on HLS-Eval~\cite{abikaram2025hlseval} with only 7B parameters, surpassing GPT-5.1 on functional pass@5, while achieving 8.5$\times$ faster training than real-reward RL. On QoR evaluation, HLS-Seek achieves the lowest latency on 19/30 kernels and Pareto-dominates HLS-specific baselines on 9 kernels.

cs.LG↗
arXiv2026

Latent Action as Intention Enables Efficient Future Imagination for World Action Models

World action models (WAMs) improve robot control by modeling how observations evolve, but generating future observations at test time incurs substantial latency. Fast-WAM removes this process for efficiency; however, our matched implementations show lower generalization for Fast-WAM than for future-aware alternatives, especially with scarce robot demonstrations and in out-of-distribution scenarios. To bridge this gap, we introduce **LAWA**, a WAM architecture that uses compact latent actions as an operational representation of future intentions, enabling efficient test-time future imagination without generating future observations. Specifically, a discrete tokenizer enhanced by action-free pre-training produces manipulation-centric codebook targets. LAWA jointly denoises a continuous latent state anchored to these targets with executable action chunks while omitting the future-video branch at inference. On RoboCasa, LAWA achieves state-of-the-art average success rates of 65.6% and 80.8% in the few-shot and full data settings, improving over the matched Fast-WAM baseline by 9.6 and 4.5 points, respectively. It also preserves the performance level of the matched Joint-WAM variant while requiring 42.9% lower inference latency. LAWA also demonstrates competitive zero-shot robustness on LIBERO-Plus and superior performance on real-world tasks. These results show that future imagination need not be discarded: retaining it with compact latent actions yields an effective trade-off among performance, generalization, and latency. Code and models will be released.

cs.RO↗
arXiv2026

TUTTI: Toward generalizable audio-to-score transcription via fully synthesized data

Generalizable Audio-to-Score (A2S) transcription is fundamentally constrained by the severe scarcity of high-quality, real-world paired data. Relying solely on existing human-annotated datasets often restricts the generalization of A2S models, limiting their efficacy primarily to single-instrumentation domains. To break this dependency on scarce real-world data, we introduce TUTTI (Transformer for Unified audio-To-score Transcription trained on Synthetic multi-Instrumentation Data), a pre-training paradigm driven by a purely synthetic, large-scale dataset. Rather than using human-composed scores, we leverage a symbolic music generation model to generate a massive, highly scalable multi-instrumentation corpus and create audio-score pairs with expressive acoustic characteristics. Capitalizing on the generated data, we employ a standard Transformer encoder-decoder architecture. We empirically demonstrate that pre-training a unified attention-based model on generated, multi-instrumentation data yields a consistently stronger foundational representation than single-instrumentation training. When fine-tuned with downstream real-world datasets, TUTTI outperforms previous approaches, establishing new overall state-of-the-art results across various A2S baselines. Notably, TUTTI shows remarkable cross-instrument transferability, effectively adapting to unseen instruments with highly competitive performance. The source code and the TuttiCorpus dataset will be made publicly available at https://github.com/a-musiclover/TUTTI.

cs.SD↗
arXiv2026

A Hierarchical Synergistic Deep Learning Framework Integrating Composition, Structure, and Ionic Transport for Solid-State Electrolyte Discovery

Inorganic solid-state electrolytes must combine high room-temperature ionic conductivity, a wide electrochemical window, excellent electronic insulation, and favorable mechanical compliance. Single models struggle to support reliable multi-objective screening across vast chemical spaces because of training-data distribution mismatch, cross-property dataset heterogeneity, and scarce kinetic transport data. To overcome these limitations, we develop a hierarchical synergistic deep-learning framework that sequentially coordinates efficiency, accuracy, and reliability through four complementary modules. The in-house-developed L-G-DCNN and a multi-fidelity implementation built on DenseGNN serve as compositional and structural experts for thermodynamic coarse screening and multi-property evaluation, respectively; MatterSim and system-specific DeePMD models provide transport pre-assessment and kinetic validation. Systematic benchmarks show that each module outperforms mainstream counterparts in its task, while retrospective validation establishes dual closed-loop verification of module-level accuracy and end-to-end workflow reliability. Applied to 30,364,908 Alex/ICSD-derived candidates, the framework identifies 97 high-performance candidates with room-temperature ionic conductivities of 0.109--59.0 mS/cm, including 94 halides, one borohydride, and two oxides. Consistency with independent experimental data confirms that 76 of the 94 halides fall within reported high-conductivity structural regions. Analysis reveals that Li$^{+}$ jump-network connectivity, rather than the number of geometric Li sites, is the core determinant of room-temperature ionic conductivity. Li-defect engineering effectively enhances oxide transport, whereas the inherent rigidity of the O$^{2-}$ framework suggests a potential upper limit on oxide electrolyte performance.

cond-mat.mtrl-sci↗
arXiv2026

DICS: Data-Informed Centroid Splitting for Decision Tree Classifiers

Decision tree-based models are widely used in machine learning due to their interpretability and strong empirical performance. However, training decision trees can be computationally expensive, particularly for large and high-dimensional datasets, largely due to the exhaustive search over candidate splits at each node. To improve computational efficiency, we propose Data-Informed Centroid Splitting (DICS), a clustering-based framework that constructs a compact and informative set of candidate splits using data-driven priors. By incorporating class-aware structure, DICS significantly reduces the split search space for classification tasks while preserving predictive performance. We further provide theoretical analysis showing that under the stated assumptions, DICS does not degrade the performance of classification trees compared to exhaustive split search. DICS can be incorporated into classification trees, random forests, and gradient-boosting models. Extensive experiments demonstrate that DICS achieves comparable accuracy while substantially reducing training time across synthetic and benchmark datasets, highlighting the benefit of integrating data-informed priors into split selection for scalable classification tree learning.

cs.LG↗
arXiv2026

Human-Guided Causal Knowledge Injection for Virtual Cells

Virtual cells employ machine learning models to simulate and predict cellular behaviors, serving as a critical computational framework for investigating health and disease. Injecting causal graphs into virtual cells can improve the interpretability, but such graphs are usually not available in real-world applications. Recently, many methods have been proposed to construct causal graphs from data, which group genes based on their similarities to form concepts and extract their causal relationships. However, since this automatic process is unsupervised, the causal graphs usually contain errors. In this paper, we propose a human-guided causal knowledge injection method for virtual cells. We developed a gene-similarity-aware causal graph visualization supported by a hybrid optimization algorithm to help explore both the causal relationships between concepts and the similarities between genes. Based on the exploration, we further developed a counterfactual analysis strategy supported by a counterfactual visualization and a causal path visualization to help validate and refine causal graphs. The effectiveness of our method is demonstrated through two real-world case studies, the extraction of scientifically meaningful causal insights, and positive feedback from domain experts.

cs.HC↗
arXiv2026

On the number of small edge-weighted subgraphs

Subgraph counting is a fundamental task that underpins several network analysis methodologies, including community detection and graph two-sample tests. Counting subgraphs is a computationally intensive problem. Substantial research has focused on developing efficient algorithms and strategies to make it feasible for larger unweighted graphs. Implementing those algorithms can be a significant hurdle for data professionals or researchers with limited expertise in algorithmic principles and programming. Furthermore, many real-world networks are weighted. Computing the number of weighted subgraphs in weighted networks presents a computational challenge, as no efficient algorithm exists for the worst-case scenario. In this paper, we derive explicit formulas for counting small edge-weighted subgraphs using the weighted adjacency matrix. These formulas are applicable to unweighted networks, offering a simple and highly practical analytical tool for researchers across various scientific domains. In addition, we introduce a generalized methodology for calculating arbitrary weighted subgraphs.

math.CO↗
arXiv2026

Quantum-Structured World Models (QSWMs) for Predictive Latent Dynamics

World models learn latent states that summarize interaction histories, evolve over time, and support prediction, simulation, or planning. Most existing world models represent these states using classical vectors, probability distributions, recurrent hidden states, or transformer activations. In this paper, we introduce Quantum-Structured World Models (QSWMs), a quantum-inspired framework for predictive world modeling with structured latent states, latent transition operators, and measurement-inspired decoding maps. We study whether mathematical structures inspired by quantum theory, such as complex-valued representations and density-matrix-like latents, provide useful inductive biases for world modeling. We establish three foundational properties: classical inclusion, predictive sufficiency, and structured compactness. We then instantiate complex-valued and density-matrix-like QSWM variants and evaluate them on elementary cellular automata against strong classical baselines. Results show promising local predictive potential for complex-valued QSWMs, while also revealing limitations in long-horizon rollout, density-matrix variants

cs.LG↗
arXiv2026

Can Large Language Models Recover Semantic Optimization Opportunities That Compilers Miss?

Optimizing compilers miss profitable transformations when their enabling semantics are absent from the analyzed program representation. We ask whether large language models (LLMs) can recover such semantics from heterogeneous C/C++ context and realize them as validated, contract-preserving artifacts. We introduce SeGaBench, an executable benchmark containing 100 synthetic and 20 source-backed cases spanning low-level assumptions, data-structure invariants, and high-level semantic lifting. Each case includes hidden enabling semantics, an oracle artifact, correctness and semantic validators, and a reproducible performance protocol. We evaluate five LLMs using five independent responses per case. The strongest model produces correct artifacts in 94.8% of responses, achieves at least 1.05x speedup in 83.3%, and obtains a performance success on 93.3% of cases. Nevertheless, correct artifacts often close only part of the oracle gap. These results show that LLMs can complement compiler analysis as speculative semantic proposers, provided that their artifacts are validated and evaluated.

cs.PL↗
arXiv2026

SymphonyGen: 3D Hierarchical Orchestral Generation with Controllable Harmony Skeleton

Generating symphonic music requires simultaneously managing high-level structural form and dense, multi-track orchestration, yet existing symbolic models often struggle with a "complexity-control imbalance" between scalability and steerability. We present SymphonyGen, a 3D hierarchical framework for contemporary orchestral generation, whose cascading decoders decompose the bar, track, and event axes, keeping decoding memory far below flat token streams and enabling conditioning at every structural level. A beat-quantized multi-pitch harmony skeleton, which may be user-written, analyzed, or model-generated, provides "short-score" conditioning, enabling outline control while producing orchestral textures. The model is refined with reinforcement learning against a cross-modal acoustic reward from CLaMP 3 audio embeddings, and a dissonance-averse sampling algorithm suppresses unintended tonal clashes during inference. Objective evaluations show that both post-training mechanisms reduce dissonance while maintaining independent melodic metrics, and in subjective tests SymphonyGen is rated above baseline systems in quality and preference, significantly so among general listeners.

cs.SD↗
arXiv2026

adabay: an R package for rapid evaluation and calibration of Bayesian group sequential designs across common endpoint types

Bayesian group sequential designs (GSDs) combine the efficiency of frequentist GSDs with clinically interpretable probability statements and principled external evidence incorporation. However, their uptake in confirmatory trials has been held back by the cost of evaluating frequentist operating characteristics at the design stage, which often nests Markov chain Monte Carlo or another approximate posterior inference within a Monte Carlo trial-simulation loop. We introduce adabay, an open-source R package implementing a semi-simulation framework for the rapid evaluation and calibration of Bayesian GSDs. Trial data paths are simulated by Monte Carlo, while per-look posteriors and posterior tail probabilities are computed analytically or by low-dimensional deterministic quadrature. Flexible prior specification is achieved by approximating any user-specified prior with a finite mixture of conjugate components, with tail-probability diagnostics that flag inadequate approximations at the decision thresholds. The package offers a unified application programming interface for continuous, binary, count and time-to-event endpoints, supports posterior-probability decision rules with one or more efficacy and futility criteria under binding or non-binding regimes, and includes a precomputation strategy that decouples threshold calibration and look-time selection from the simulation pass. adabay reproduces the operating characteristics of BATSS and adaptr in the continuous and binary case studies, the analytic gsbDesign values in the continuous case, and the BATSS results in the count case, all within Monte Carlo error, while running approximately four to five orders of magnitude faster than BATSS and one to over two orders of magnitude faster than adaptr per virtual trial on eight cores. The package is distributed under an MIT licence.

stat.ME↗
arXiv2026

BMOA: Baseline-Mechanism-Outcome Attribution for Compiler-Induced Numerical Deviations

Formalizing compiler-aware numerical correctness requires distinguishing what an observed floating-point difference means, what compiler behavior the evidence supports, and what numerical consequence follows. Existing testing workflows often collapse these questions into a pass/fail mismatch. We introduce Baseline--Mechanism--Outcome Attribution (BMOA), a diagnostic framework that separates the comparison relation and system boundary, the evidence-supported compiler mechanism, and the reference-qualified accuracy outcome. BMOA combines operational strict floating-point, transformation-local, reproducibility, cross-compiler, and higher-precision comparisons, while preserving mixed, ambiguous, and unknown attributions when evidence is insufficient. Each record retains inputs, configurations, numerical metrics, and supporting artifacts for audit. We evaluate BMOA on six scientific-computing kernels, deterministic stress-input families, and controlled Clang configurations on ARM64. A 1,276-record attribution corpus and a 162-instance controlled mechanism matrix show that baseline choice changes diagnoses, compiler-induced deviation does not imply accuracy loss, and cancellation and large dynamic range expose the strongest effects within the targeted matrix. BMOA converts raw mismatches into explicit, auditable, evidence-bounded records. Although it is not itself a proof system, these records provide an empirical foundation for future formal specifications and proof obligations for compiler-aware numerical correctness.

cs.PL↗