arXiv subjects

Sijie Ren

Publications and source records attributed to Sijie Ren.

13 recordsLinked to original sources

arXiv2026

CardioBench: A Real-World Data Benchmark for Evaluating Large Language Models in Clinically Authentic Cardiovascular Care Scenarios

Background: Most medical large language model (LLM) benchmarks focus on examination knowledge or isolated tasks and may not reflect the longitudinal, multimodal, and safety-critical workflow of cardiovascular care. Objective: To develop CardioBench, a real-world benchmark spanning the cardiovascular care continuum, and assess LLM performance across clinical dimensions and specialist tasks. Methods: CardioBench includes 2,263 items from 13 task-specific datasets derived from de-identified cardiovascular records and examination data. Sixteen cardiology physicians conducted annotation and reference construction, followed by cross-review from two senior cardiologists. Seven LLMs generated 15,841 outputs under standardized zero-shot settings. Open-ended tasks were evaluated using key-point coverage and holistic clinical quality, while CardioEthics was scored by accuracy. Results: GPT-5.4 achieved the highest macro-average (62.55) and item-weighted mean (62.19), followed by Gemini 3.1 Pro (59.95) and Qwen 3.6 27B (59.72). GPT-5.4 ranked first in all three dimensions. CardioAuxReport performed best (86.38), whereas CardioECGRead (17.25) and CardioEthics (17.34) were lowest. The largest gaps between holistic clinical quality and key-point coverage occurred in CardioComm (52.71), CardioEmergRescue (52.05), and CardioTreatPlan (48.80). Conclusions: To our knowledge, CardioBench is the largest real-world, multi-task benchmark for LLM evaluation across the cardiovascular care continuum and offers the broadest coverage of clinically authentic cardiology scenarios reported to date. It provides a rigorous framework for identifying model strengths, clinically important omissions, and priorities for future development.

cs.CL↗

arXiv2026

MedBench v5: A Dynamic, Process-Oriented, and Hallucination-Aware Benchmark for Clinical Multimodal Models

Existing medical AI benchmarks lack process visibility, atomic skill evaluation, and integrated hallucination detection. We introduce MedBench v5, a redesigned benchmark for clinical multimodal models (language, vision-language, and agent systems) that moves from static QA to dynamic, process-oriented evaluation. MedBench v5 features: (1) a dual-dimensional framework combining Clinical Cognitive Responsiveness (13 sub-dimensions) and Medical Atomic Skills (4 agent environments), covering 63 tasks; (2) three switchable information-flow stressors (omission, contradiction, evidence delay) for factorized degradation analysis; (3) a dynamic process audit protocol with five reasoning nodes that produces model-specific failure fingerprints; (4) hallucination propagation monitoring across initiation, propagation, anchoring, and contradiction interaction-capturing silent hallucination. Experiments on frontier models show that strong overall task performance does not guarantee process stability: stressors mainly disrupt contradiction detection, diagnosis updating, hallucination propagation, and contradiction-based self-correction, while final evidence grounding can remain superficially stable. MedBench v5 provides a unified infrastructure for capability profiling, controllable stress testing, process auditing, and hallucination trajectory analysis in clinical AI evaluation.

cs.CL↗

arXiv2025

Andr{á}sfai--Erdős--Sós theorem under max-degree constraints

We establish the following strengthening of the celebrated Andr{á}sfai--Erdős--Sós theorem: If $G$ is an $n$-vertex $K_{r+1}$-free graph whose minimum degree $δ(G)$ and maximum degree $Δ(G)$ satisfy \begin{align*} δ(G) > \min \left\{ \frac{3r-4}{3r-2}n-\frac{Δ(G)}{3r-2},~n-\frac{Δ(G)+1}{r-1} \right\}, \end{align*} then $G$ is $r$-partite. This bound is tight for all feasible values of $Δ(G)$. We also obtain an analogous tight result for graphs with large odd girth. Our proof does not rely on the Andr{á}sfai--Erdős--Sós theorem itself, and therefore yields an alternative proof of this classical result.

math.CO↗

arXiv2025

MedBench v4: A Robust and Scalable Benchmark for Evaluating Chinese Medical Language Models, Multimodal Models, and Intelligent Agents

Recent advances in medical large language models (LLMs), multimodal models, and agents demand evaluation frameworks that reflect real clinical workflows and safety constraints. We present MedBench v4, a nationwide, cloud-based benchmarking infrastructure comprising over 700,000 expert-curated tasks spanning 24 primary and 91 secondary specialties, with dedicated tracks for LLMs, multimodal models, and agents. Items undergo multi-stage refinement and multi-round review by clinicians from more than 500 institutions, and open-ended responses are scored by an LLM-as-a-judge calibrated to human ratings. We evaluate 15 frontier models. Base LLMs reach a mean overall score of 54.1/100 (best: Claude Sonnet 4.5, 62.5/100), but safety and ethics remain low (18.4/100). Multimodal models perform worse overall (mean 47.5/100; best: GPT-5, 54.9/100), with solid perception yet weaker cross-modal reasoning. Agents built on the same backbones substantially improve end-to-end performance (mean 79.8/100), with Claude Sonnet 4.5-based agents achieving up to 85.3/100 overall and 88.9/100 on safety tasks. MedBench v4 thus reveals persisting gaps in multimodal reasoning and safety for base models, while showing that governance-aware agentic orchestration can markedly enhance benchmarked clinical readiness without sacrificing capability. By aligning tasks with Chinese clinical guidelines and regulatory priorities, the platform offers a practical reference for hospitals, developers, and policymakers auditing medical AI.

cs.CL↗

arXiv2025

MedCalc-Eval and MedCalc-Env: Advancing Medical Calculation Capabilities of Large Language Models

As large language models (LLMs) enter the medical domain, most benchmarks evaluate them on question answering or descriptive reasoning, overlooking quantitative reasoning critical to clinical decision-making. Existing datasets like MedCalc-Bench cover few calculation tasks and fail to reflect real-world computational scenarios. We introduce MedCalc-Eval, the largest benchmark for assessing LLMs' medical calculation abilities, comprising 700+ tasks across two types: equation-based (e.g., Cockcroft-Gault, BMI, BSA) and rule-based scoring systems (e.g., Apgar, Glasgow Coma Scale). These tasks span diverse specialties including internal medicine, surgery, pediatrics, and cardiology, offering a broader and more challenging evaluation setting. To improve performance, we further develop MedCalc-Env, a reinforcement learning environment built on the InternBootcamp framework, enabling multi-step clinical reasoning and planning. Fine-tuning a Qwen2.5-32B model within this environment achieves state-of-the-art results on MedCalc-Eval, with notable gains in numerical sensitivity, formula selection, and reasoning robustness. Remaining challenges include unit conversion, multi-condition logic, and contextual understanding. Code and datasets are available at https://github.com/maokangkun/MedCalc-Eval.

cs.CL↗

arXiv2025

Extremal triangle-free graphs with chromatic number at least four

Let $G$ be an $n$-vertex triangle-free graph. The celebrated Mantel's theorem showed that $e(G)\leq \lfloor\frac{n^2}{4}\rfloor$. In 1962, Erdős (together with Gallai), and independently Andrásfai, proved that if $G$ is non-bipartite then $e(G)\leq \lfloor\frac{(n-1)^2}{4}\rfloor+1$. In this paper, we extend this result and show that if $G$ has chromatic number at least four and $n\geq 90$, then $e(G)\leq \lfloor\frac{(n-3)^2}{4}\rfloor+5$. The blow-ups of Grötzsch graph shows that this bound is best possible.

math.CO↗

arXiv2024

Positive codegree Andrásfai--Erdős--Sós theorem for the generalized triangle

The celebrated Andrásfai--Erdős--Sós Theorem from 1974 shows that every $n$-vertex triangle-free graph with minimum degree greater than $2n/5$ must be bipartite. We establish a positive codegree extension of this result for the $r$-uniform generalized triangle $\mathrm{T}_{r} = \left\{\{1,\ldots, r-1,r\}, \{1,\ldots, r-1,r+1\},\{r,r+1, \ldots, 2r-1\}\right\}$$\colon$ For every $n \ge (r-1)(2r+1)/2$, if $\mathcal{H}$ is an $n$-vertex $\mathrm{T}_{r}$-free $r$-uniform hypergraph in which each $(r-1)$-tuple of vertices is contained in either zero edges or more than $2n/(2r+1)$ edges of $\mathcal{H}$, then $\mathcal{H}$ is $r$-partite. This result provides the first tight positive codegree Andr{á}sfai--Erdős--Sós type theorem for hypergraphs. It also immediately implies that the positive codegree Turán number of $\mathrm{T}_{r}$ is $\lfloor n/r \rfloor$ for all $r$. Additionally, for $r=3$, our result answers one of the questions posed by Hou et al.~\cite{HLYZZ22} in a strong form.

math.CO↗

arXiv2024

Andrásfai--Erdős--Sós theorem for the generalized triangle

The celebrated Andrásfai--Erdős--Sós Theorem from 1974 shows that every $n$-vertex triangle-free graph with minimum degree greater than $2n/5$ must be bipartite. Its extensions to $3$-uniform hypergraphs without the generalized triangle $F_5 = \{abc, abd, cde\}$ have been explored in several previous works such as~\cite{LMR23unif,HLZ24}, demonstrating the existence of $\varepsilon > 0$ such that for large $n$, every $n$-vertex $F_5$-free $3$-graph with minimum degree greater than $(1/9-\varepsilon) n^2$ must be $3$-partite. We determine the optimal value for $\varepsilon$ by showing that for $n \ge 5000$, every $n$-vertex $F_5$-free $3$-graph with minimum degree greater than $4n^2/45$ must be $3$-partite, thus establishing the first tight Andrásfai--Erdős--Sós type theorem for hypergraphs. As a corollary, for all positive $n$, every $n$-vertex cancellative $3$-graph with minimum degree greater than $4n^2/45$ must be $3$-partite. This result is also optimal and considerably strengthens prior work, such as that by Bollobás~\cite{Bol74} and Keevash--Mubayi~\cite{KM04Cancel}.

math.CO↗

arXiv2024

Enhancing the Reliability of Segment Anything Model for Auto-Prompting Medical Image Segmentation with Uncertainty Rectification

The Segment Anything Model (SAM) has recently emerged as a groundbreaking foundation model for prompt-driven image segmentation tasks. However, both the original SAM and its medical variants require slice-by-slice manual prompting of target structures, which directly increase the burden for applications. Despite attempts of auto-prompting to turn SAM into a fully automatic manner, it still exhibits subpar performance and lacks of reliability especially in the field of medical imaging. In this paper, we propose UR-SAM, an uncertainty rectified SAM framework to enhance the reliability for auto-prompting medical image segmentation. Building upon a localization framework for automatic prompt generation, our method incorporates a prompt augmentation module to obtain a series of input prompts for SAM for uncertainty estimation and an uncertainty-based rectification module to further utilize the distribution of estimated uncertainty to improve the segmentation performance. Extensive experiments on two public 3D medical datasets covering the segmentation of 35 organs demonstrate that without supplementary training or fine-tuning, our method further improves the segmentation performance with up to 10.7 % and 13.8 % in dice similarity coefficient, demonstrating efficiency and broad capabilities for medical image segmentation without manual prompting.

cs.CV↗

arXiv2023

A stability result for $C_{2k+1}$-free graphs

A graph $G$ is called $C_{2k+1}$-free if it does not contain any cycle of length $2k+1$. In 1981, Haggkvist, Faudree and Schelp showed that every $n$-vertex triangle-free graph with more than $\frac{(n-1)^2}{4}+1$ edges is bipartite. In this paper, we extend their result and show that for $1\leq t\leq 2k-2$ and $n\geq 318t^2k$, every $n$-vertex $C_{2k+1}$-free graph with more than $\frac{(n-t-1)^2}{4}+\binom{t+2}{2}$ edges can be made bipartite by either deleting at most $t-1$ vertices or deleting at most $\binom{\lfloor\frac{t+2}{2}\rfloor}{2}+\binom{\lceil\frac{t+2}{2}\rceil}{2}-1$ edges. The construction shows that this is best possible.

math.CO↗

arXiv2021

Forecasting Irreversible Disease via Progression Learning

Forecasting Parapapillary atrophy (PPA), i.e., a symptom related to most irreversible eye diseases, provides an alarm for implementing an intervention to slow down the disease progression at early stage. A key question for this forecast is: how to fully utilize the historical data (e.g., retinal image) up to the current stage for future disease prediction? In this paper, we provide an answer with a novel framework, namely \textbf{D}isease \textbf{F}orecast via \textbf{P}rogression \textbf{L}earning (\textbf{DFPL}), which exploits the irreversibility prior (i.e., cannot be reversed once diagnosed). Specifically, based on this prior, we decompose two factors that contribute to the prediction of the future disease: i) the current disease label given the data (retinal image, clinical attributes) at present and ii) the future disease label given the progression of the retinal images that from the current to the future. To model these two factors, we introduce the current and progression predictors in DFPL, respectively. In order to account for the degree of progression of the disease, we propose a temporal generative model to accurately generate the future image and compare it with the current one to get a residual image. The generative model is implemented by a recurrent neural network, in order to exploit the dependency of the historical data. To verify our approach, we apply it to a PPA in-house dataset and it yields a significant improvement (\textit{e.g.}, \textbf{4.48\%} of accuracy; \textbf{3.45\%} of AUC) over others. Besides, our generative model can accurately localize the disease-related regions.

cs.CV↗

arXiv2020

A Holistically-Guided Decoder for Deep Representation Learning with Applications to Semantic Segmentation and Object Detection

Both high-level and high-resolution feature representations are of great importance in various visual understanding tasks. To acquire high-resolution feature maps with high-level semantic information, one common strategy is to adopt dilated convolutions in the backbone networks to extract high-resolution feature maps, such as the dilatedFCN-based methods for semantic segmentation. However, due to many convolution operations are conducted on the high-resolution feature maps, such methods have large computational complexity and memory consumption. In this paper, we propose one novel holistically-guided decoder which is introduced to obtain the high-resolution semantic-rich feature maps via the multi-scale features from the encoder. The decoding is achieved via novel holistic codeword generation and codeword assembly operations, which take advantages of both the high-level and low-level features from the encoder features. With the proposed holistically-guided decoder, we implement the EfficientFCN architecture for semantic segmentation and HGD-FPN for object detection and instance segmentation. The EfficientFCN achieves comparable or even better performance than state-of-the-art methods with only 1/3 of their computational costs for semantic segmentation on PASCAL Context, PASCAL VOC, ADE20K datasets. Meanwhile, the proposed HGD-FPN achieves $>2\%$ higher mean Average Precision (mAP) when integrated into several object detection frameworks with ResNet-50 encoding backbones.

cs.CV↗

arXiv2020

AIM 2020 Challenge on Learned Image Signal Processing Pipeline

This paper reviews the second AIM learned ISP challenge and provides the description of the proposed solutions and results. The participating teams were solving a real-world RAW-to-RGB mapping problem, where to goal was to map the original low-quality RAW images captured by the Huawei P20 device to the same photos obtained with the Canon 5D DSLR camera. The considered task embraced a number of complex computer vision subtasks, such as image demosaicing, denoising, white balancing, color and contrast correction, demoireing, etc. The target metric used in this challenge combined fidelity scores (PSNR and SSIM) with solutions' perceptual results measured in a user study. The proposed solutions significantly improved the baseline results, defining the state-of-the-art for practical image signal processing pipeline modeling.

cs.CV↗