arXiv subjects

Ao Sun

Publications and source records attributed to Ao Sun.

At least 55 records · Page 3Linked to original sources

arXiv2025

Biomarker combination based on the Youden index with and without gold standard

In clinical practice, multiple biomarkers are often measured on the same subject for disease diagnosis, and combining them can improve diagnostic accuracy. Existing studies typically combine multiple biomarkers by maximizing the Area Under the ROC Curve (AUC), assuming a gold standard exists or that biomarkers follow a multivariate normal distribution. However, practical diagnostic settings require both optimal combination coefficients and an effective cutoff value, and the reference test may be imperfect. In this paper, we propose a two-stage method for identifying the optimal linear combination and cutoff value based on the Youden index. First, it maximizes an approximation of the empirical AUC to estimate the optimal linear coefficients for combining multiple biomarkers. Then, it maximizes the empirical Youden index to determine the optimal cutoff point for disease classification. Under the semiparametric single index model and regularity conditions, the estimators for the linear coefficients, cutoff point, and Youden index are consistent. This method is also applicable when the reference standard is imperfect. We demonstrate the performance of our method through simulations and apply it to construct a diagnostic scale for Chinese medicine.

stat.ME↗

arXiv2025

NavigateDiff: Visual Predictors are Zero-Shot Navigation Assistants

Navigating unfamiliar environments presents significant challenges for household robots, requiring the ability to recognize and reason about novel decoration and layout. Existing reinforcement learning methods cannot be directly transferred to new environments, as they typically rely on extensive mapping and exploration, leading to time-consuming and inefficient. To address these challenges, we try to transfer the logical knowledge and the generalization ability of pre-trained foundation models to zero-shot navigation. By integrating a large vision-language model with a diffusion network, our approach named \mname ~constructs a visual predictor that continuously predicts the agent's potential observations in the next step which can assist robots generate robust actions. Furthermore, to adapt the temporal property of navigation, we introduce temporal historical information to ensure that the predicted image is aligned with the navigation scene. We then carefully designed an information fusion framework that embeds the predicted future frames as guidance into goal-reaching policy to solve downstream image navigation tasks. This approach enhances navigation control and generalization across both simulated and real-world environments. Through extensive experimentation, we demonstrate the robustness and versatility of our method, showcasing its potential to improve the efficiency and effectiveness of robotic navigation in diverse settings.

cs.RO↗

arXiv2025

Passing through nondegenerate singularities in mean curvature flows

In this paper, we study the properties of nondegenerate cylindrical singularities of mean curvature flow. We prove they are isolated in spacetime and provide a complete description of the geometry and topology change of the flow passing through the singularities. Particularly, the topology change agrees with the level sets change near a critical point of a Morse function, which is the same as performing surgery. The proof is based on a new $L^2$-distance monotonicity formula, which allows us to derive a discrete almost monotonicity of the ``decay order", a discrete mean curvature flow analog to Almgren's frequency function.

math.DG↗

arXiv2025

CHAIR -- Classifier of Hallucination as Improver

In this work, we introduce CHAIR (Classifier of Hallucination As ImproveR), a supervised framework for detecting hallucinations by analyzing internal logits from each layer of every token. Our method extracts a compact set of features such as maximum, minimum, mean, standard deviation, and slope-from the token logits across all layers, enabling effective hallucination detection without overfitting. Experiments on TruthfulQA and MMLU datasets demonstrate that CHAIR significantly improves detection accuracy, particularly in zero-shot scenarios, showcasing its robustness and generalizability. Beyond hallucination detection, CHAIR highlights the potential of using internal representations for designing advanced decoding strategies. By leveraging patterns in logits, we suggest that more sophisticated models and adaptive decoding methods could further reduce hallucinations and enhance text completion quality. CHAIR not only offers a practical solution for detecting hallucinations but also lays the groundwork for exploring richer representations in LLMs to improve their factuality and coherence.

cs.CL↗

arXiv2024

How Multi-Modal LLMs Reshape Visual Deep Learning Testing? A Comprehensive Study Through the Lens of Image Mutation

Visual deep learning (VDL) systems have shown significant success in real-world applications like image recognition, object detection, and autonomous driving. To evaluate the reliability of VDL, a mainstream approach is software testing, which requires diverse mutations over image semantics. The rapid development of multi-modal large language models (MLLMs) has introduced revolutionary image mutation potentials through instruction-driven methods. Users can now freely describe desired mutations and let MLLMs generate the mutated images. Hence, parallel to large language models' (LLMs) recent success in traditional software fuzzing, one may also expect MLLMs to be promising for VDL testing in terms of offering unified, diverse, and complex image mutations. However, the quality and applicability of MLLM-based mutations in VDL testing remain largely unexplored. We present the first study, aiming to assess MLLMs' adequacy from 1) the semantic validity of MLLM mutated images, 2) the alignment of MLLM mutated images with their text instructions (prompts), and 3) the faithfulness of how different mutations preserve semantics that are ought to remain unchanged. With large-scale human studies and quantitative evaluations, we identify MLLM's promising potentials in expanding the covered semantics of image mutations. Notably, while SoTA MLLMs (e.g., GPT-4V) fail to support or perform worse in editing existing semantics in images (as in traditional mutations like rotation), they generate high-quality test inputs using "semantic-replacement" mutations (e.g., "dress a dog with clothes"), which bring extra semantics to images; these were infeasible for past approaches. Hence, we view MLLM-based mutations as a vital complement to traditional mutations, and advocate future VDL testing tasks to combine MLLM-based methods and traditional image mutations for comprehensive and reliable testing.

cs.SE↗

arXiv2024

Seq1F1B: Efficient Sequence-Level Pipeline Parallelism for Large Language Model Training

The emergence of large language models (LLMs) relies heavily on distributed training strategies, among which pipeline parallelism plays a crucial role. As LLMs' training sequence length extends to 32k or even 128k, the current pipeline parallel methods face severe bottlenecks, including high memory footprints and substantial pipeline bubbles, greatly hindering model scalability and training throughput. To enhance memory efficiency and training throughput, in this work, we introduce an efficient sequence-level one-forward-one-backward (1F1B) pipeline scheduling method tailored for training LLMs on long sequences named Seq1F1B. Seq1F1B decomposes batch-level schedulable units into finer sequence-level units, reducing bubble size and memory footprint. Considering that Seq1F1B may produce slight extra bubbles if sequences are split evenly, we design a computation-wise strategy to partition input sequences and mitigate this side effect. Compared to competitive pipeline baseline methods such as Megatron 1F1B pipeline parallelism, our method achieves higher training throughput with less memory footprint. Notably, Seq1F1B efficiently trains a LLM with 30B parameters on sequences up to 64k using 64 NVIDIA A100 GPUs without recomputation strategies, a feat unachievable with existing methods. Our source code is based on Megatron-LM, and now is avaiable at: https://github.com/MayDomine/Seq1F1B.git.

cs.DC↗

arXiv2024

SDI-Net: Toward Sufficient Dual-View Interaction for Low-light Stereo Image Enhancement

Currently, most low-light image enhancement methods only consider information from a single view, neglecting the correlation between cross-view information. Therefore, the enhancement results produced by these methods are often unsatisfactory. In this context, there have been efforts to develop methods specifically for low-light stereo image enhancement. These methods take into account the cross-view disparities and enable interaction between the left and right views, leading to improved performance. However, these methods still do not fully exploit the interaction between left and right view information. To address this issue, we propose a model called Toward Sufficient Dual-View Interaction for Low-light Stereo Image Enhancement (SDI-Net). The backbone structure of SDI-Net is two encoder-decoder pairs, which are used to learn the mapping function from low-light images to normal-light images. Among the encoders and the decoders, we design a module named Cross-View Sufficient Interaction Module (CSIM), aiming to fully exploit the correlations between the binocular views via the attention mechanism. The quantitative and visual results on public datasets validate the superiority of our method over other related methods. Ablation studies also demonstrate the effectiveness of the key elements in our model.

cs.CV↗

arXiv2024

Integrated high-performance error correction for continuous-variable quantum key distribution

An integrated error-correction scheme with high throughput, low frame errors rate (FER) and high reconciliation efficiency under low signal to noise ratio (SNR) is one of the major bottlenecks to realize high-performance and low-cost continuous variable quantum key distribution (CV-QKD). To solve this long-standing problem, a novel two-stage error correction method with limited precision that is suitable for integration given limited on-chip hardware resource while maintaining excellent decoding performance is proposed, and experimentally verified on a commercial FPGA. Compared to state-of-art results, the error-correction throughput can be improved more than one order of magnitude given FER<0.1 based on the proposed method, where 544.03 Mbps and 393.33 Mbps real-time error correction is achieved for typical 0.2 and 0.1 code rate, respectively. Besides, compared with traditional decoding method, the secure key rate (SKR) for CV-QKD under composable security framework can be improved by 140.09% and 122.03% by using the proposed two-stage decoding method for codes rate 0.2 and 0.1, which can support 32.70 Mbps and 5.66 Mbps real-time SKR under typical transmission distances of 25 km and 50 km, correspondingly. The record-breaking results paves the way for large-scale deployment of high-rate integrated CV-QKD systems in metropolitan quantum secure network.

quant-ph↗

arXiv2024

Secure Combination of Untrusted Time information Based on Optimized Dempster-Shafer Theory

Secure precision time synchronization is important for applications of Cyber-Physical Systems. However, several attacks, especially the Time Delay Attack (TDA), deteriorates the performance of time synchronization system seriously. Multiple paths scheme is thought as an effective security countermeasure to decrease the influence of TDA. However, the effective secure combination algorithm is still missed for precision time synchronization. In this paper, a secure combination algorithm based on Dempster-Shafer theory is proposed for multiple paths method. Special optimizations are done for the combination algorithm to solve the potential problems due to untrusted evidence. Theoretical simulation shows that the proposed algorithm works much better than Fault Tolerant Algorithm (FTA) and the attack detection method based on single path. And experimental demonstration proves the feasibility and superiority of the proposed algorithm, where the time stability with 27.97 ps, 1.57 ps, and 1.12 ps at average time 1s, 10s, 100s is achieved under TDA and local clock jump. The proposed algorithm can be used to improve the security and resilience of many importance synchronization protocol, such as NTP, PTP, and TWFTT.

cs.CR↗

arXiv2024

BurstAttention: An Efficient Distributed Attention Framework for Extremely Long Sequences

Effective attention modules have played a crucial role in the success of Transformer-based large language models (LLMs), but the quadratic time and memory complexities of these attention modules also pose a challenge when processing long sequences. One potential solution for the long sequence problem is to utilize distributed clusters to parallelize the computation of attention modules across multiple devices (e.g., GPUs). However, adopting a distributed approach inevitably introduces extra memory overheads to store local attention results and incurs additional communication costs to aggregate local results into global ones. In this paper, we propose a distributed attention framework named ``BurstAttention'' to optimize memory access and communication operations at both the global cluster and local device levels. In our experiments, we compare BurstAttention with other competitive distributed attention solutions for long sequence processing. The experimental results under different length settings demonstrate that BurstAttention offers significant advantages for processing long sequences compared with these competitive baselines, reducing 40% communication overheads and achieving 1.37 X speedup during training 128K sequence length on 32 X A100.

cs.DC↗

arXiv2024

Existence of minimal hypersurfaces with arbitrarily large area and possible obstructions

We prove that in a closed Riemannian manifold with dimension between $3$ and $7$, either there are minimal hypersurfaces with arbitrarily large area, or there exist uncountably many stable minimal hypersurfaces. Moreover, the latter case has a very pathological Cantor set structure which does not show up in certain manifolds. Among the applications, we prove that there exist minimal hypersurfaces with arbitrarily large area in analytic manifolds. In the proof, we use the Almgren-Pitts min-max theory proposed by Marques-Neves, the ideas developed by Song in his proof of Yau's conjecture, and the resolution of the generic multiplicity-one conjecture by Zhou.

math.DG↗

arXiv2024

On the probability of a Pareto record

Given a sequence of independent random vectors taking values in ${\mathbb R}^d$ and having common continuous distribution function $F$, say that the $n^{\rm \scriptsize th}$ observation sets a (Pareto) record if it is not dominated (in every coordinate) by any preceding observation. Let $p_n(F) \equiv p_{n, d}(F)$ denote the probability that the $n^{\rm \scriptsize th}$ observation sets a record. There are many interesting questions to address concerning $p_n$ and multivariate records more generally, but this short paper focuses on how $p_n$ varies with $F$, particularly if, under $F$, the coordinates exhibit negative dependence or positive dependence (rather than independence, a more-studied case). We introduce new notions of negative and positive dependence ideally suited for such a study, called negative record-setting probability dependence (NRPD) and positive record-setting probability dependence (PRPD), relate these notions to existing notions of dependence, and for fixed $d \geq 2$ and $n \geq 1$ prove that the image of the mapping $p_n$ on the domain of NRPD (respectively, PRPD) distributions is $[p^*_n, 1]$ (resp., $[n^{-1}, p^*_n]$), where $p^*_n$ is the record-setting probability for any continuous $F$ governing independent coordinates.

math.PR↗

arXiv2024

Generic regularity of Level Set Flows with spherical singularity

The sphere is well-known as the only generic compact shrinker for mean curvature flow (MCF). In this paper, we characterize the generic dynamics of MCFs with a spherical singularity. In terms of the level set flow formulation of MCF, we establish that generically the arrival time function of level set flow with spherical singularity has at most $C^2$ regularity.

math.DG↗

arXiv2024

Sharpened localization of the trailing point of the Pareto record frontier

For $d\ge2$ and iid $d$-dimensional observations $X^{(1)},X^{(2)},\dots$ with independent Exponential$(1)$ coordinates, we revisit the study by Fill and Naiman (Electron. J. Probab., 2020) of the boundary (relative to the closed positive orthant), or "frontier", $F_n$ of the closed Pareto record-setting (RS) region \[ \mbox{RS}_n:=\{0\le x\in{\mathbb R}^d:x\not\prec X^{(i)}\mbox{\ for all $1\le i\le n$}\} \] at time $n$, where $0\le x$ means that $0\le x_j$ for $1\le j\le d$ and $x\prec y$ means that $x_j 0$ and $c_n\to\infty$ we have \[ {\mathbb P}(F_n^- -\ln n\in (-(2+\varepsilon)\ln\ln\ln n,c_n))\to 1 \] (describing typical behavior) and almost surely \[ \limsup \frac{F_n^- - \ln n}{\ln \ln n} \le 0 \quad \mbox{and} \quad \liminf \frac{F_n^- - \ln n}{\ln \ln \ln n} \in [-2, -1]. \] In this paper we use the theory of generators (minima of $F_n$) together with the first- and second-moment methods to improve considerably the trailing-point location results to \[ F_n^- - (\ln n - \ln \ln \ln n) \overset{\mathrm{P}}{\longrightarrow} - \ln(d - 1) \] (describing typical behavior) and, for $d \ge 3$, almost surely \begin{align*} &\limsup [F_n^- - (\ln n - \ln \ln \ln n)] \leq -\ln(d - 2) + \ln 2 \\ \mbox{and }&\liminf [F_n^- - (\ln n - \ln \ln \ln n)] \ge - \ln d - \ln 2. \end{align*}

math.PR↗

arXiv2024

Eliminating Information Leakage in Hard Concept Bottleneck Models with Supervised, Hierarchical Concept Learning

Concept Bottleneck Models (CBMs) aim to deliver interpretable and interventionable predictions by bridging features and labels with human-understandable concepts. While recent CBMs show promising potential, they suffer from information leakage, where unintended information beyond the concepts (either when concepts are represented with probabilities or binary states) are leaked to the subsequent label prediction. Consequently, distinct classes are falsely classified via indistinguishable concepts, undermining the interpretation and intervention of CBMs. This paper alleviates the information leakage issue by introducing label supervision in concept predication and constructing a hierarchical concept set. Accordingly, we propose a new paradigm of CBMs, namely SupCBM, which achieves label predication via predicted concepts and a deliberately-designed intervention matrix. SupCBM focuses on concepts that are mostly relevant to the predicted label and only distinguishes classes when different concepts are presented. Our evaluations show that SupCBM outperforms SOTA CBMs over diverse datasets. It also manifests better generality across different backbone models. With proper quantification of information leakage in different CBMs, we demonstrate that SupCBM significantly reduces the information leakage.

cs.LG↗

arXiv2023

An adaptable JSON Diff Framework

In this paper, we present an implementation of JSON-diff framework JYCM, extending the existing framework by introducing the concept of "unordered" comparisons and allowing users to customize their comparison scenarios flexibly. Furthermore, we provide a diff-result renderer to visualize better and understand the differences between JSON objects. Our work enables more adaptable and comprehensive comparisons to accommodate a wider range of use cases and requirements.

cs.SE↗

arXiv2023

Explain Any Concept: Segment Anything Meets Concept-Based Explanation

EXplainable AI (XAI) is an essential topic to improve human understanding of deep neural networks (DNNs) given their black-box internals. For computer vision tasks, mainstream pixel-based XAI methods explain DNN decisions by identifying important pixels, and emerging concept-based XAI explore forming explanations with concepts (e.g., a head in an image). However, pixels are generally hard to interpret and sensitive to the imprecision of XAI methods, whereas "concepts" in prior works require human annotation or are limited to pre-defined concept sets. On the other hand, driven by large-scale pre-training, Segment Anything Model (SAM) has been demonstrated as a powerful and promotable framework for performing precise and comprehensive instance segmentation, enabling automatic preparation of concept sets from a given image. This paper for the first time explores using SAM to augment concept-based XAI. We offer an effective and flexible concept-based explanation method, namely Explain Any Concept (EAC), which explains DNN decisions with any concept. While SAM is highly effective and offers an "out-of-the-box" instance segmentation, it is costly when being integrated into defacto XAI pipelines. We thus propose a lightweight per-input equivalent (PIE) scheme, enabling efficient explanation with a surrogate model. Our evaluation over two popular datasets (ImageNet and COCO) illustrate the highly encouraging performance of EAC over commonly-used XAI methods.

cs.CV↗

arXiv2023

"Oops, Did I Just Say That?" Testing and Repairing Unethical Suggestions of Large Language Models with Suggest-Critique-Reflect Process

As the popularity of large language models (LLMs) soars across various applications, ensuring their alignment with human values has become a paramount concern. In particular, given that LLMs have great potential to serve as general-purpose AI assistants in daily life, their subtly unethical suggestions become a serious and real concern. Tackling the challenge of automatically testing and repairing unethical suggestions is thus demanding. This paper introduces the first framework for testing and repairing unethical suggestions made by LLMs. We first propose ETHICSSUITE, a test suite that presents complex, contextualized, and realistic moral scenarios to test LLMs. We then propose a novel suggest-critic-reflect (SCR) process, serving as an automated test oracle to detect unethical suggestions. We recast deciding if LLMs yield unethical suggestions (a hard problem; often requiring human expertise and costly to decide) into a PCR task that can be automatically checked for violation. Moreover, we propose a novel on-the-fly (OTF) repairing scheme that repairs unethical suggestions made by LLMs in real-time. The OTF scheme is applicable to LLMs in a black-box API setting with moderate cost. With ETHICSSUITE, our study on seven popular LLMs (e.g., ChatGPT, GPT-4) uncovers in total 109,824 unethical suggestions. We apply our OTF scheme on two LLMs (Llama-13B and ChatGPT), which generates valid repair to a considerable amount of unethical ones, paving the way for more ethically conscious LLMs.

cs.SE↗