arXiv subjects

Jiang Li

Publications and source records attributed to Jiang Li.

At least 19 recordsLinked to original sources

arXiv2026

Do Large Language Models Perform Well on Comprehending Poetic Logic in Modern Chinese Poetry?

Large Language Models (LLMs) have achieved significant progress across a wide range of natural language processing (NLP) tasks, yet their ability to understand literary texts, particularly modern Chinese poetry, remains largely unexplored. The unique literary characteristics of modern Chinese poetry necessitate a distinct form of reasoning for effective comprehension. Unlike conventional texts that convey clear information, the unique "poetic logic" of modern Chinese poetry requires a holistic reasoning approach that goes beyond superficial semantic analysis to be understood. However, current evaluation paradigms largely ignore this critical dimension. To address this gap, we propose Peony, the first benchmark specifically designed for evaluating the poetic logic of modern Chinese poetry. We define poetic logic as four tasks across three levels, namely stanza, line, and imagery, and systematically evaluate and analyze six mainstream LLMs based on Peony. We evaluate these models under both non-thinking and thinking configurations. The experimental results reveal the limitations of current LLMs in understanding the poetic logic of modern Chinese poetry and validate the effectiveness and necessity of Peony. Our data and code will be available.

cs.CL↗

arXiv2026

Quantization Degradation in Large Language Models: A Signal-Noise Perspective

Post-training quantization reduces the deployment cost of large language models, yet how severely a quantized model degrades is not determined by bit-width alone. We systematically study weight-only post-training quantization across bit-widths, quantization methods, model scales and downstream tasks on multiple model families. We observe that such degradation varies substantially across these factors: 4-bit quantization usually preserves performance, 2-bit often causes broad degradation, and at 3-bit, degradation becomes apparent but varies markedly with task type, quantization method and model scale. To explain this variability, we use the signal-to-noise ratio (SNR) to measure how strongly quantization perturbs full-precision representations. We trace degradation back to two linked processes: how quantization errors arise within individual modules, and how they accumulate across layers. First, a source SNR decomposition shows that newly introduced errors depend on three factors: the magnitude of the weight error, the strength of the task-specific signal, and how strongly the quantization error aligns with task-specific activations. Different factors affect these components in distinct ways. Second, a cross-layer propagation analysis shows that these errors can be attenuated, preserved, or amplified as they pass across layers, and that larger models benefit from weaker error amplification. Together, these results establish that quantization degradation is governed by how errors are introduced at the source and how they accumulate across the network.

cs.AI↗

arXiv2026

A Heuristic Perspective on Debiasing Language Models

Language models (LMs) often acquire various biases during pre-training and may express them in interactions, potentially causing social harm. Existing methods often rely on counterfactual augmentation or representation projection. These strategies remain limited in practice due to their high computational costs and difficulty in scaling to larger models. Additionally, many of these strategies require manual data annotation, narrowing their scope to specific cultures and bias categories. To overcome these limitations, we propose HEIMAT, a HEurIstic-style autoMATic debiasing framework for LMs. HEIMAT consists of two main steps: bias disclosure and debiasing fine-tuning. In the first step, it uses simple templates to construct heuristic prompts, which are applied to reveal model biases and generate corresponding context prompts. In the second step, it fine-tunes the model by minimizing the Jensen-Shannon divergence of predictions on these context prompts to reduce bias. Extensive experiments show that HEIMAT effectively mitigates bias in different cultures while maintaining the model's natural language understanding (NLU) performance.

cs.CL↗

arXiv2026

WearWow: Native 2K Multi-Garment Virtual Try-On via Adaptive Token Packing and Preference Alignment

Synthesizing native 2K multi-garment virtual try-on is a formidable frontier in digital fashion, critically bottlenecked by two fundamental limitations: the O(N^2) memory explosion induced by 2k conditions, and the spectral bias of diffusion models that over-smooths high-frequency fabric details. We present WearWow, an end-to-end, mask-free generative framework that pioneers ultra-high-resolution multi-garment synthesis. To mitigate the memory explosion , we propose Adaptive 2D Token Packing (ATP). ATP leverages inherent garment sparsity to algorithmically pack heterogeneous items onto a unified 2D canvas and prune uninformative background tokens, minimizing the effective sequence length and subsequent memory overhead while rigorously preserving 2D spatial priors. To rectify texture degradation, we introduce the Multi-dimensional Try-on Reward (MTR) system. MTR synergizes a Semantic Guidance Reward to explicitly drive tactile restoration with a Cloth Distribution Reward to implicitly anchor the physical distribution, a joint formulation that effectively mitigates the severe reward hacking. Furthermore, we curate WearWow-2K, an extreme-quality dataset comprising native 2K triplets, providing physically correct spatial interactions that naturally empower the model's mask-free generation. Extensive experiments demonstrate that WearWow establishes a new state-of-the-art, exceeding existing commercial baselines in native 2K multi-garment synthesis.

cs.CV↗

arXiv2026

Extreme-Scale Atomistic Simulation of Real-Temperature Magnetic Skyrmion Dynamics by Coupled Spin-Lattice Modeling

Real-temperature topological magnetic dynamics in functional materials is governed by coupled lattice and spin evolution, yet remains inaccessible to predictive simulation at device-relevant scales. As a flagship example, thermally driven helix-to-skyrmion transformation in FeGe requires atomistic resolution, explicit lattice motion, and micrometer-scale domains to resolve device-scale topological texture formation. We combine a spin-constrained density-functional-theory-trained neuro-evolution potential with a structure-preserving spin-lattice integrator within one machine-learned framework. Architecture-specific optimizations, kernel fusion, SVE2 vectorization, and NUMA-aware data layout deliver a seven orders-of-magnitude speedup over prior spin-aware methods. Deployed on LineShine exascale supercomputer, the full application scales to 12.45 million CPU cores with 89.7% weak-scaling efficiency, enabling simulations of 1.34 trillion atoms and an equal number of spins while reaching 48.5 PFLOPS in double precision. The simulations directly resolve real-temperature skyrmion nucleation and reorganization at previously inaccessible scales, establishing a new regime for predictive simulation of coupled spin-lattice topological magnetic dynamics.

cs.DC↗

arXiv2026

A Unified and Reproducible Experimentation Framework for Speech Understanding

Speech foundation models and Speech LLMs have advanced speech understanding, yet deployment-oriented model selection is hindered by non-comparable evaluations caused by mismatched post-processing, and by training results that are hard to reproduce across data scales and pipelines. We present SURE, a unified experimentation framework that standardizes prediction formats, normalization, and scoring. SURE evaluates strong systems across paradigms, from conventional pipelines to Speech LLMs, on representative tasks under realistic acoustic and linguistic stressors. Beyond evaluation, SURE introduces an agent-assisted training conversion flow that maps paper and code into versioned, runnable training pipelines under a unified protocol on matched open-data subsets. Overall, SURE improves comparability and reproducibility for deployment-oriented evaluation.

eess.AS↗

arXiv2026

AtelierEval: Agentic Evaluation of Humans & LLMs as Text-to-Image Prompters

Text-to-image (T2I) systems increasingly rely on upstream prompters, either humans or multimodal large language models (MLLMs), to translate user intent into detailed prompts. Yet current benchmarks fix the prompt and only evaluate T2I models, leaving the prompting proficiency of this upstream component entirely unmeasured. We introduce AtelierEval, the first unified benchmark that quantifies prompting proficiency across 360 expert-crafted tasks. Grounded in a cognitive view, it spans three task categories and instantiates tasks using a taxonomy of real-world challenges, with a dual interface for both humans and MLLMs. To enable scalable and reliable evaluation, we propose AtelierJudge, a skill-based, memory-augmented agentic evaluator. It produces subjective and objective scores for prompt-image pairs, achieving a Spearman correlation of 0.79 with human experts, approaching human performance. Extensive experiments benchmark 8 MLLMs against 48 human users across 4 T2I backends, validate AtelierEval as a robust diagnostic tool, and reveal the superiority of mimicry over planning, advocating for an image-augmented direction for future prompters. Our work is released to support future research.

cs.AI↗

arXiv2026

Training-Inference Consistent Segmented Execution for Long-Context LLMs

Transformer-based large language models face severe scalability challenges in long-context generation due to the computational and memory costs of full-context attention. Under practical computation and memory constraints, many inference-efficient long-context methods improve efficiency by adopting bounded-context or segment-level execution only during inference, while continuing to train models under full-context attention, resulting in a mismatch between training and inference execution and state-transition semantics. Based on this insight, we propose a training-inference consistent segment-level generation framework, in which training and inference follow the same segment-level forward execution semantics. During training, consistency with inference is enforced by restricting gradient propagation to KV states carried over from the immediately preceding segment, while permitting head-specific access to past KV states during the forward pass without involving them in gradient propagation. Across long-context benchmarks, our approach achieves performance comparable to full-context attention, while achieving competitive latency-memory trade-offs against strong inference-efficient baselines, and substantially improving scalability at very long context lengths (e.g., approximately 6x lower peak prefill memory at 128K compared to full-context attention with FlashAttention).

cs.CL↗

arXiv2026

Boxing inequalities for relative fractional perimeter and fractional Poincar\'e-type inequalities on John domains with the BBM factor

For $0<\delta,\tau<1$ and $1\le s\le \frac{n}{n-\delta}$, we prove that for a given $s$-John domain $\Omega\subset \mathbb{R}^n$, the following Boxing inequality holds for every Lebesgue measurable set $U\subset\Omega$ with $|U|/|\Omega|\le\gamma<1$: \[ \mathcal{H}^{s(n-\delta)}_{\infty}(U\setminus\mathcal{N}_U)\le C(1-\delta)\int_\Omega\int_{|x-y|<\tau\operatorname{dist}(y,\partial\Omega)}\frac{|\chi_U(x)-\chi_U(y)|}{|x-y|^{n+\delta}}\,dx\,dy, \] where $\mathcal{H}^{s(n-\delta)}_{\infty}(U)$ denotes the $s(n-\delta)$-dimensional Hausdorff content of $U$, $\mathcal{N}_U$ is a set of Lebesgue measure zero and the constant $C$ depends only on $n,\tau,s,\gamma$, the John constant and the diameter of $\Omega$. Moreover, we establish the functional formulation of the above Boxing inequality and discuss the equivalence between these two formulations. Based on the Boxing inequality, we prove the fractional Poincar\'e--Wirtinger trace inequality on $s$-John domains, of which the fractional Sobolev--Poincar\'e inequality and fractional Hardy-type inequality are special cases. Notably, we prove all of the aforementioned inequalities with the Bourgain--Brezis--Mironescu (BBM) factor $1-\delta$. Furthermore, with the aid of the Bourgain--Brezis--Mironescu formula, we recover the Poincar\'e--Wirtinger trace inequality. Finally, by showing that, under the separation property, any domain supporting the Boxing inequality is necessarily a John domain, we conclude that the John domain condition is essentially sharp for the above inequalities. All the above inequalities with the BBM factor are new even for Lipschitz domains.

math.FA↗

arXiv2026

From Signal Degradation to Computation Collapse: Uncovering the Two Failure Modes of LLM Quantization

Post-Training Quantization (PTQ) is critical for the efficient deployment of Large Language Models (LLMs). While 4-bit quantization is widely regarded as an optimal trade-off, reducing the precision to 2-bit usually triggers a catastrophic ``performance cliff.'' It remains unclear whether the underlying mechanisms differ fundamentally. Consequently, we conduct a systematic mechanistic analysis, revealing two qualitatively distinct failure modes: Signal Degradation, where the computational patterns remain intact but information precision is impaired by cumulative error; and Computation Collapse, where key components fail to function, preventing correct information processing and destroying the signal in the early layers. Guided by this diagnosis, we conduct mechanism-aware interventions, demonstrating that targeted, training-free repair can mitigate Signal Degradation, but remains ineffective for Computation Collapse. Our findings provide a systematic diagnostic framework for PTQ failures and suggest that addressing Computation Collapse requires structural reconstruction rather than mere compensation.

cs.CL↗

arXiv2026

Exploring the Capability Boundaries of LLMs in Mastering of Chinese Chouxiang Language

While large language models (LLMs) have achieved remarkable success in general language tasks, their performance on Chouxiang Language, a representative subcultural language in the Chinese internet context, remains largely unexplored. In this paper, we introduce Mouse, a specialized benchmark designed to evaluate the capabilities of LLMs on NLP tasks involving Chouxiang Language across six tasks. Experimental results show that, current state-of-the-art (SOTA) LLMs exhibit clear limitations on multiple tasks, while performing well on tasks that involve contextual semantic understanding. In addition, we further discuss the reasons behind the generally low performance of SOTA LLMs on Chouxiang Language, examine whether the LLM-as-a-judge approach adopted for translation tasks aligns with human judgments and values, and analyze the key factors that influence Chouxiang translation. Our study aims to promote further research in the NLP community on multicultural integration and the dynamics of evolving internet languages. Our code and data are publicly available.

cs.CL↗

arXiv2026

Who Wrote This Line? Evaluating the Detection of LLM-Generated Classical Chinese Poetry

The rapid development of large language models (LLMs) has extended text generation tasks into the literary domain. However, AI-generated literary creations has raised increasingly prominent issues of creative authenticity and ethics in literary world, making the detection of LLM-generated literary texts essential and urgent. While previous works have made significant progress in detecting AI-generated text, it has yet to address classical Chinese poetry. Due to the unique linguistic features of classical Chinese poetry, such as strict metrical regularity, a shared system of poetic imagery, and flexible syntax, distinguishing whether a poem is authored by AI presents a substantial challenge. To address these issues, we introduce ChangAn, a benchmark for detecting LLM-generated classical Chinese poetry that containing total 30,664 poems, 10,276 are human-written poems and 20,388 poems are generated by four popular LLMs. Based on ChangAn, we conducted a systematic evaluation of 12 AI detectors, investigating their performance variations across different text granularities and generation strategies. Our findings highlight the limitations of current Chinese text detectors, which fail to serve as reliable tools for detecting LLM-generated classical Chinese poetry. These results validate the effectiveness and necessity of our proposed ChangAn benchmark. Our dataset and code are available at https://github.com/VelikayaScarlet/ChangAn.

cs.CL↗

arXiv2025

Geometric inequalities related to fractional perimeter: fractional Poincar\'e, isoperimetric, and boxing inequalities in metric measure spaces

In the setting of a complete, doubling metric measure space $(X,d,\mu)$ supporting a $(1,1)$-Poincar\'e inequality, we show that for all $0<\theta<1$, the following fractional Poincar\'e inequality holds for all balls $B$ and locally integrable functions $u$, $$ \int_{B}|u-u_B|d\mu\le C(1-\theta)\,\text{rad}(B)^\theta\int_{\tau B}\int_{\tau B}\frac{|u(x)-u(y)|}{d(x,y)^\theta\mu(B(x,d(x,y)))}d\mu(y)d\mu(x), $$ where $C\ge 1$ and $\tau\ge 1$ are constants depending only on the doubling and $(1,1)$-Poincar\'e inequality constants. Notably, this inequality features the scaling constant $(1-\theta)$ present in the Bourgain-Brezis-Mironescu theory characterizing Sobolev functions via nonlocal functionals. From this inequality, we obtain a fractional relative isoperimetric inequality as well as global and local versions of a fractional boxing inequality, each featuring the same scaling constant $(1-\theta)$ and defined in terms of the fractional $\theta$-perimeter, and prove equivalences with the above fractional Poincar\'e inequality. We also show that $(X,d,\mu)$ supports a $(1,1)$-Poincar\'e inequality if and only if the above fractional Poincar\'e inequality holds for all $\theta$ sufficiently close to $1$. Under the additional assumption of lower Ahlfors $Q$-regularity of the measure $\mu$, we additionally use the aforementioned results to establish global inequalities, in the form of fractional isoperimetric and fractional Sobolev inequalities, which also feature the scaling constant $(1-\theta)$. Moreover, we prove that such inequalities are equivalent with the lower Ahlfors $Q$-regularity condition on the measure.

math.FA↗

arXiv2025

Task-Stratified Knowledge Scaling Laws for Post-Training Quantized Large Language Models

Post-Training Quantization (PTQ) is a critical strategy for efficient Large Language Models (LLMs) deployment. However, existing scaling laws primarily focus on general performance, overlooking crucial fine-grained factors and how quantization differentially impacts diverse knowledge capabilities. To address this, we establish Task-Stratified Knowledge Scaling Laws. By stratifying capabilities into memorization, application, and reasoning, we develop a framework that unifies model size, bit-width, and fine-grained factors: group size and calibration set size. Validated on 293 diverse PTQ configurations, our framework demonstrates strong fit and cross-architecture consistency. It reveals distinct sensitivities across knowledge capabilities: reasoning is precision-critical, application is scale-responsive, and memorization is calibration-sensitive. We highlight that in low-bit scenarios, optimizing these fine-grained factors is essential for preventing performance collapse. These findings provide an empirically-backed foundation for designing knowledge-aware quantization strategies.

cs.CL↗

arXiv2025

Automated Multi-label Classification of Eleven Retinal Diseases: A Benchmark of Modern Architectures and a Meta-Ensemble on a Large Synthetic Dataset

The development of multi-label deep learning models for retinal disease classification is often hindered by the scarcity of large, expertly annotated clinical datasets due to patient privacy concerns and high costs. The recent release of SynFundus-1M, a high-fidelity synthetic dataset with over one million fundus images, presents a novel opportunity to overcome these barriers. To establish a foundational performance benchmark for this new resource, we developed an end-to-end deep learning pipeline, training six modern architectures (ConvNeXtV2, SwinV2, ViT, ResNet, EfficientNetV2, and the RETFound foundation model) to classify eleven retinal diseases using a 5-fold multi-label stratified cross-validation strategy. We further developed a meta-ensemble model by stacking the out-of-fold predictions with an XGBoost classifier. Our final ensemble model achieved the highest performance on the internal validation set, with a macro-average Area Under the Receiver Operating Characteristic Curve (AUC) of 0.9973. Critically, the models demonstrated strong generalization to three diverse, real-world clinical datasets, achieving an AUC of 0.7972 on a combined DR dataset, an AUC of 0.9126 on the AIROGS glaucoma dataset and a macro-AUC of 0.8800 on the multi-label RFMiD dataset. This work provides a robust baseline for future research on large-scale synthetic datasets and establishes that models trained exclusively on synthetic data can accurately classify multiple pathologies and generalize effectively to real clinical images, offering a viable pathway to accelerate the development of comprehensive AI systems in ophthalmology.

cs.CV↗

arXiv2025

The Role of AI in Facilitating Interdisciplinary Collaboration: Evidence from AlphaFold

The acceleration of artificial intelligence (AI) in science is recognized and many scholars have begun to explore its role in interdisciplinary collaboration. However, the mechanisms and extent of this impact are still unclear. This study, using AlphaFold's impact on structural biologists, examines how AI technologies influence interdisciplinary collaborative patterns. By analyzing 1,247 AlphaFold-related papers and 7,700 authors from Scopus, we employ bibliometric analysis and causal inference to compare interdisciplinary collaboration between AlphaFold adopters and non-adopters. Contrary to the widespread belief that AI facilitates interdisciplinary collaboration, our findings show that AlphaFold increased structural biology-computer science collaborations by just 0.48%, with no measurable effect on other disciplines. Specifically, AI creates interdisciplinary collaboration demands with specific disciplines due to its technical characteristics, but this demand is weakened by technological democratization and other factors. These findings demonstrate that artificial intelligence (AI) alone has limited efficacy in bridging disciplinary divides or fostering meaningful interdisciplinary collaboration.

cs.DL↗

arXiv2025

McBE: A Multi-task Chinese Bias Evaluation Benchmark for Large Language Models

As large language models (LLMs) are increasingly applied to various NLP tasks, their inherent biases are gradually disclosed. Therefore, measuring biases in LLMs is crucial to mitigate its ethical risks. However, most existing bias evaluation datasets focus on English and North American culture, and their bias categories are not fully applicable to other cultures. The datasets grounded in the Chinese language and culture are scarce. More importantly, these datasets usually only support single evaluation tasks and cannot evaluate the bias from multiple aspects in LLMs. To address these issues, we present a Multi-task Chinese Bias Evaluation Benchmark (McBE) that includes 4,077 bias evaluation instances, covering 12 single bias categories, 82 subcategories and introducing 5 evaluation tasks, providing extensive category coverage, content diversity, and measuring comprehensiveness. Additionally, we evaluate several popular LLMs from different series and with parameter sizes. In general, all these LLMs demonstrated varying degrees of bias. We conduct an in-depth analysis of results, offering novel insights into bias in LLMs.

cs.CL↗

arXiv2025

Degradation-Aware Image Enhancement via Vision-Language Classification

Image degradation is a prevalent issue in various real-world applications, affecting visual quality and downstream processing tasks. In this study, we propose a novel framework that employs a Vision-Language Model (VLM) to automatically classify degraded images into predefined categories. The VLM categorizes an input image into one of four degradation types: (A) super-resolution degradation (including noise, blur, and JPEG compression), (B) reflection artifacts, (C) motion blur, or (D) no visible degradation (high-quality image). Once classified, images assigned to categories A, B, or C undergo targeted restoration using dedicated models tailored for each specific degradation type. The final output is a restored image with improved visual quality. Experimental results demonstrate the effectiveness of our approach in accurately classifying image degradations and enhancing image quality through specialized restoration models. Our method presents a scalable and automated solution for real-world image enhancement tasks, leveraging the capabilities of VLMs in conjunction with state-of-the-art restoration techniques.

cs.CV↗