arXiv subjects

Jinfeng Li

Publications and source records attributed to Jinfeng Li.

At least 19 recordsLinked to original sources

arXiv2026

The Note-Chord-Voice Framework: Structured Source Separation and Causal Inference for EV Charging Data

Real-world EV charging data exhibit three interlocking pathologies: hardware fragmentation (network timeouts and billing resets split sessions), physical violations (independent energy/duration models produce impossible states like 50 kWh in 10 min on a 7 kW charger), and collider bias (clustering on post-treatment outcomes opens backdoor paths for price elasticity). We propose the Note-Chord-Voice framework, a music-inspired, axiom-driven pipeline that separates data cleaning (Repair Chords), structural pattern discovery (Harmonic Chords), descriptive source separation (NMF Voices), and causal inference into distinct, falsifiable stages. Key innovations: (i) falsification gates (A1-A5, G3, G10) that test data suitability before modeling; (ii) Gamma-initialized NMF with input rescaling for convergence stability from STL decomposition; (iii) tag-based coupon grading (A/B/C/D) to isolate quasi-random treatment from night-time confounders and targeted promotions; (iv) separate per-voice OLS to avoid simplex collinearity; (v) Foote novelty curves for structural regime detection. Applied to the Jiangmen dataset (495,707 sessions, 20 stations, from July 2024 to March 2025), all core axioms pass except G3 (no strong 168 h cycle). NMF achieves R^2=0.9921; the physically constrained duration model yields aggregate R^2=0.5409. Two voices are price-sensitive (beta = -11 to -14 min, p<0.001), of which one is stable (Voice 3, beta=-14.16) and one treatment-driven (Voice 1, beta=-11.10); only the stable voice supports causal claims. Counterfactual simulation shows targeting discounts to price-sensitive voices recovers 52.8% of discount expenditures (~0.85M CNY/year); restricting to the single stable price-sensitive voice yields a more conservative estimate.

eess.SP↗

arXiv2026

Auditing Chinese Web-scale Corpora via Sampled BPE Token Statistics

Chinese web pollution has surfaced in LLMs, motivating audits of upstream Chinese corpora. However, auditing such corpora faces three challenges: (1) their web-scale size makes full scan costly; (2) prior analyses are often too coarse to expose token-level pollution; (3) Chinese web pollution is implicit and rapidly changing. We propose Sampled-BPE, a lightweight token-level auditing pipeline that sample a small subset and train BPE tokenizer to surface polluted tokens. Experiments show that Sampled-BPE preserves usable estimates while substantially reducing runtime and memory: a 148.4 $\times$ speedup and a 35.8 $\times$ memory reduction induce only 4.25% relative error for pollution categories. We apply the pipeline to 11 open Chinese corpora and 6 Chinese Common Crawl snapshots from 2021 to 2026. The audit reveals widespread but uneven pollution across open corpora, as well as highly polluted and temporally shifting Chinese web content. We further release a hierarchical Chinese web token dataset with 660k+ token records, each with web context, category, and explanation fields, organized as trees to support review and tracing of pollution.

cs.CL↗

arXiv2026

Can Released LLM Vocabularies Support Token-Level Estimation of Hidden Corpora?

Pretraining corpus composition shapes LLM capabilities, but it often remains hidden even when model weights are released. Prior work has inferred corpus mixtures or traced specific token groups from released tokenizer vocabularies; in contrast, we estimate corpus ratios for arbitrary target tokens. We first show that BPE tokenizers trained on different corpora share stable token ID--ratio distributions, motivating distribution transfer from known corpora to a target tokenizer trained on hidden corpora. We then propose Quantile-Guided Density Estimation (QGDE), which approximates this distribution with multiple quantile trends and uses local density weighting to produce token-level estimates. In controlled settings and a realistic setting using the released SmolLM tokenizer, QGDE achieves mean relative errors as low as 3.00% for token-level estimation and 3.08% after aggregation into category-level mixtures. These results suggest that released tokenizer vocabularies provide a useful signal for fine-grained corpus estimation beyond coarse composition inference.

cs.CL↗

arXiv2026

YuFeng-XGuard: A Reasoning-Centric, Interpretable, and Flexible Guardrail Model for Large Language Models

As large language models (LLMs) are increasingly deployed in real-world applications, safety guardrails are required to go beyond coarse-grained filtering and support fine-grained, interpretable, and adaptable risk assessment. However, existing solutions often rely on rapid classification schemes or post-hoc rules, resulting in limited transparency, inflexible policies, or prohibitive inference costs. To this end, we present YuFeng-XGuard, a reasoning-centric guardrail model family designed to perform multi-dimensional risk perception for LLM interactions. Instead of producing opaque binary judgments, YuFeng-XGuard generates structured risk predictions, including explicit risk categories and configurable confidence scores, accompanied by natural language explanations that expose the underlying reasoning process. This formulation enables safety decisions that are both actionable and interpretable. To balance decision latency and explanatory depth, we adopt a tiered inference paradigm that performs an initial risk decision based on the first decoded token, while preserving ondemand explanatory reasoning when required. In addition, we introduce a dynamic policy mechanism that decouples risk perception from policy enforcement, allowing safety policies to be adjusted without model retraining. Extensive experiments on a diverse set of public safety benchmarks demonstrate that YuFeng-XGuard achieves stateof-the-art performance while maintaining strong efficiency-efficacy trade-offs. We release YuFeng-XGuard as an open model family, including both a full-capacity variant and a lightweight version, to support a wide range of deployment scenarios.

cs.CL↗

arXiv2025

MIRAGE: Misleading Retrieval-Augmented Generation via Black-box and Query-agnostic Poisoning Attacks

Retrieval-Augmented Generation (RAG) systems enhance LLMs with external knowledge but introduce a critical attack surface: corpus poisoning. While recent studies have demonstrated the potential of such attacks, they typically rely on impractical assumptions, such as white-box access or known user queries, thereby underestimating the difficulty of real-world exploitation. In this paper, we bridge this gap by proposing MIRAGE, a novel multi-stage poisoning pipeline designed for strict black-box and query-agnostic environments. Operating on surrogate model feedback, MIRAGE functions as an automated optimization framework that integrates three key mechanisms: it utilizes persona-driven query synthesis to approximate latent user search distributions, employs semantic anchoring to imperceptibly embed these intents for high retrieval visibility, and leverages an adversarial variant of Test-Time Preference Optimization (TPO) to maximize persuasion. To rigorously evaluate this threat, we construct a new benchmark derived from three long-form, domain-specific datasets. Extensive experiments demonstrate that MIRAGE significantly outperforms existing baselines in both attack efficacy and stealthiness, exhibiting remarkable transferability across diverse retriever-LLM configurations and highlighting the urgent need for robust defense strategies.

cs.CR↗

arXiv2025

Enhancing New-item Fairness in Dynamic Recommender Systems

New-items play a crucial role in recommender systems (RSs) for delivering fresh and engaging user experiences. However, traditional methods struggle to effectively recommend new-items due to their short exposure time and limited interaction records, especially in dynamic recommender systems (DRSs) where new-items get continuously introduced and users' preferences evolve over time. This leads to significant unfairness towards new-items, which could accumulate over the successive model updates, ultimately compromising the stability of the entire system. Therefore, we propose FairAgent, a reinforcement learning (RL)-based new-item fairness enhancement framework specifically designed for DRSs. It leverages knowledge distillation to extract collaborative signals from traditional models, retaining strong recommendation capabilities for old-items. In addition, FairAgent introduces a novel reward mechanism for recommendation tailored to the characteristics of DRSs, which consists of three components: 1) a new-item exploration reward to promote the exposure of dynamically introduced new-items, 2) a fairness reward to adapt to users' personalized fairness requirements for new-items, and 3) an accuracy reward which leverages users' dynamic feedback to enhance recommendation accuracy. Extensive experiments on three public datasets and backbone models demonstrate the superior performance of FairAgent. The results present that FairAgent can effectively boost new-item exposure, achieve personalized new-item fairness, while maintaining high recommendation accuracy.

cs.IR↗

arXiv2024

fairBERTs: Erasing Sensitive Information Through Semantic and Fairness-aware Perturbations

Pre-trained language models (PLMs) have revolutionized both the natural language processing research and applications. However, stereotypical biases (e.g., gender and racial discrimination) encoded in PLMs have raised negative ethical implications for PLMs, which critically limits their broader applications. To address the aforementioned unfairness issues, we present fairBERTs, a general framework for learning fair fine-tuned BERT series models by erasing the protected sensitive information via semantic and fairness-aware perturbations generated by a generative adversarial network. Through extensive qualitative and quantitative experiments on two real-world tasks, we demonstrate the great superiority of fairBERTs in mitigating unfairness while maintaining the model utility. We also verify the feasibility of transferring adversarial components in fairBERTs to other conventionally trained BERT-like models for yielding fairness improvements. Our findings may shed light on further research on building fairer fine-tuned PLMs.

cs.CL↗

arXiv2024

S-Eval: Towards Automated and Comprehensive Safety Evaluation for Large Language Models

Generative large language models (LLMs) have revolutionized natural language processing with their transformative and emergent capabilities. However, recent evidence indicates that LLMs can produce harmful content that violates social norms, raising significant concerns regarding the safety and ethical ramifications of deploying these advanced models. Thus, it is both critical and imperative to perform a rigorous and comprehensive safety evaluation of LLMs before deployment. Despite this need, owing to the extensiveness of LLM generation space, it still lacks a unified and standardized risk taxonomy to systematically reflect the LLM content safety, as well as automated safety assessment techniques to explore the potential risk efficiently. To bridge the striking gap, we propose S-Eval, a novel LLM-based automated Safety Evaluation framework with a newly defined comprehensive risk taxonomy. S-Eval incorporates two key components, i.e., an expert testing LLM ${M}_t$ and a novel safety critique LLM ${M}_c$. ${M}_t$ is responsible for automatically generating test cases in accordance with the proposed risk taxonomy. ${M}_c$ can provide quantitative and explainable safety evaluations for better risk awareness of LLMs. In contrast to prior works, S-Eval is efficient and effective in test generation and safety evaluation. Moreover, S-Eval can be flexibly configured and adapted to the rapid evolution of LLMs and accompanying new safety threats, test generation methods and safety critique methods thanks to the LLM-based architecture. S-Eval has been deployed in our industrial partner for the automated safety evaluation of multiple LLMs serving millions of users, demonstrating its effectiveness in real-world scenarios. Our benchmark is publicly available at https://github.com/IS2Lab/S-Eval.

cs.CR↗

arXiv2024

On the equivalence of two spinodal decomposition criteria with a case study of Fe${}_{15}$Co${}_{15}$Ni${}_{35}$Cu${}_{35}$ multicomponent alloy

Spinodal decomposition in multicomponent alloys has attracted increasing attention due to its beneficial effect on their mechanical and functional properties and potential applications. Both based on the Cahn-Hillard equation, the reference element method (REM) and the projection matrix method (PMM) are the two main methods to predict the occurrence of spinodal decomposition in multicomponent alloys. In this work, it is mathematically proven that the two methods are equivalent, and therefore the advanced results based on one method can be applied to the other. Based on these methods, the $Fe{}_{15}$Co${}_{15}$Ni${}_{35}$Cu${}_{35}$ multicomponent alloy is designed as a case study. Experimental results confirm the spinodal decomposition in the heat-treated alloy, and its strength and ductility are simultaneously enhanced. This work can be the pavement for further theoretical and experimental studies on the spinodal decomposition in multicomponent alloys.

cond-mat.mtrl-sci↗

arXiv2023

A Note Related to Graph Theory

This article foucuses on $(P_3\cup P_2,K_4)$-free graph. In this paper, we prove that if G is $(P_3\cup P_2,K_4)$-free, then $χ(G)\le 7$. We then use our result to obtain the upper bound of order and chromatic number of $(4K_1,\overline{P_3\cup P_2},K_ω)$-free graph .

math.CO↗

arXiv2023

On the chromatic number of some ($P_3\cup P_2$)-free graphs

A hereditary class $\cal G$ of graphs is {\em $χ$-bounded} if there is a {\em $χ$-binding function}, say $f$, such that $χ(G)\le f(ω(G))$ for every $G\in\cal G$, where $χ(G)(ω(G))$ denotes the chromatic (clique) number of $G$. It is known that for every $(P_3\cup P_2)$-free graph $G$, $χ(G)\le \frac{1}{6}ω(G)(ω(G)+1)(ω(G)+2)$ \cite{BA18}, and the class of $(2K_2, 3K_1)$-free graphs does not admit a linear $χ$-binding function\cite{BBS19}. In this paper, we prove that (\romannumeral 1) $χ(G)\le2ω(G)$ if $G$ is ($P_3\cup P_2$, kite)-free, (\romannumeral 2) $χ(G)\leω^2(G)$ if $G$ is ($P_3\cup P_2$, hammer)-free, (\romannumeral 3) $χ(G)\le\frac{3ω^2(G)+ω(G)}{2}$ if $G$ is ($P_3\cup P_2, C_5$)-free. Furthermore, we also discuss $χ$-binding functions for $(P_3\cup P_2, K_4)$-free graphs.

math.CO↗

arXiv2023

A tight linear chromatic bound for ($P_3\cup P_2, W_4$)-free graphs

For two vertex disjoint graphs $H$ and $F$, we use $H\cup F$ to denote the graph with vertex set $V(H)\cup V(F)$ and edge set $E(H)\cup E(F)$, and use $H+F$ to denote the graph with vertex set $V(H)\cup V(F)$ and edge set $E(H)\cup E(F)\cup\{xy\;|\; x\in V(H), y\in V(F)$$\}$. A $W_4$ is the graph $K_1+C_4$. In this paper, we prove that $χ(G)\le 2ω(G)$ if $G$ is a ($P_3\cup P_2, W_4$)-free graph. This bound is tight when $ω=2$ and $3$, and improves the main result of Wang and Zhang. Also, this bound partially generalizes some results of Prashant {\em et al.}.

math.CO↗

arXiv2023

Optimal chromatic bound for ($P_3\cup P_2$, house)-free graphs

Let $G$ and $H$ be two vertex disjoint graphs. The {\em union} $G\cup H$ is the graph with $V(G\cup H)=V(G)\cup V(H)$ and $E(G\cup H)=E(G)\cup E(H)$. We use $P_k$ to denote a {\em path} on $k$ vertices, use {\em house} to denote the complement of $P_5$. In this paper, we show that $χ(G)\le2ω(G)$ if $G$ is ($P_3\cup P_2$, house)-free. Moreover, this bound is optimal when $ω(G)\ge2$.

math.CO↗

arXiv2023

Ultra-small topological spin textures with size of 1.3nm at above room temperature in Fe78Si9B13 amorphous alloy

Topologically protected spin textures, such as skyrmions1,2 and vortices3,4, are robust against perturbations, serving as the building blocks for a range of topological devices5-9. In order to implement these topological devices, it is necessary to find ultra-small topological spin textures at room temperature, because small size implies the higher topological charge density, stronger signal of topological transport10,11 and the higher memory density or integration for topological quantum devices5-9. However, finding ultra-small topological spin textures at high temperatures is still a great challenge up to now. Here we find ultra-small topological spin textures in Fe78Si9B13 amorphous alloy. We measured a large topological Hall effect (THE) up to above room temperature, indicating the existence of highly densed and ultra-small topological spin textures in the samples. Further measurements by small-angle neutron scattering (SANS) reveal that the average size of ultra-small magnetic texture is around 1.3nm. Our Monte Carlo simulations show that such ultra-small spin texture is topologically equivalent to skyrmions, which originate from competing frustration and Dzyaloshinskii-Moriya interaction12,13 coming from amorphous structure14-17. Taking a single topological spin texture as one bit and ignoring the distance between them, we evaluated the ideal memory density of Fe78Si9B13, which reaches up to 4.44*104 gigabits (43.4 TB) per in2 and is 2 times of the value of GdRu2Si218 at 5K. More important, such high memory density can be obtained at above room temperature, which is 4 orders of magnitude larger than the value of other materials at the same temperature. These findings provide a unique candidate for magnetic memory devices with ultra-high density.

cond-mat.mtrl-sci↗

arXiv2023

FairRec: Fairness Testing for Deep Recommender Systems

Deep learning-based recommender systems (DRSs) are increasingly and widely deployed in the industry, which brings significant convenience to people's daily life in different ways. However, recommender systems are also shown to suffer from multiple issues,e.g., the echo chamber and the Matthew effect, of which the notation of "fairness" plays a core role.While many fairness notations and corresponding fairness testing approaches have been developed for traditional deep classification models, they are essentially hardly applicable to DRSs. One major difficulty is that there still lacks a systematic understanding and mapping between the existing fairness notations and the diverse testing requirements for deep recommender systems, not to mention further testing or debugging activities. To address the gap, we propose FairRec, a unified framework that supports fairness testing of DRSs from multiple customized perspectives, e.g., model utility, item diversity, item popularity, etc. We also propose a novel, efficient search-based testing approach to tackle the new challenge, i.e., double-ended discrete particle swarm optimization (DPSO) algorithm, to effectively search for hidden fairness issues in the form of certain disadvantaged groups from a vast number of candidate groups. Given the testing report, by adopting a simple re-ranking mitigation strategy on these identified disadvantaged groups, we show that the fairness of DRSs can be significantly improved. We conducted extensive experiments on multiple industry-level DRSs adopted by leading companies. The results confirm that FairRec is effective and efficient in identifying the deeply hidden fairness issues, e.g., achieving 95% testing accuracy with half to 1/8 time.

cs.AI↗

arXiv2022

CAINNFlow: Convolutional block Attention modules and Invertible Neural Networks Flow for anomaly detection and localization tasks

Detection of object anomalies is crucial in industrial processes, but unsupervised anomaly detection and localization is particularly important due to the difficulty of obtaining a large number of defective samples and the unpredictable types of anomalies in real life. Among the existing unsupervised anomaly detection and localization methods, the NF-based scheme has achieved better results. However, the two subnets (complex functions) $s_{i}(u_{i})$ and $t_{i}(u_{i})$ in NF are usually multilayer perceptrons, which need to squeeze the input visual features from 2D flattening to 1D, destroying the spatial location relationship in the feature map and losing the spatial structure information. In order to retain and effectively extract spatial structure information, we design in this study a complex function model with alternating CBAM embedded in a stacked $3\times3$ full convolution, which is able to retain and effectively extract spatial structure information in the normalized flow model. Extensive experimental results on the MVTec AD dataset show that CAINNFlow achieves advanced levels of accuracy and inference efficiency based on CNN and Transformer backbone networks as feature extractors, and CAINNFlow achieves a pixel-level AUC of $98.64\%$ for anomaly detection in MVTec AD.

cs.CV↗

arXiv2022

RoChBert: Towards Robust BERT Fine-tuning for Chinese

Despite of the superb performance on a wide range of tasks, pre-trained language models (e.g., BERT) have been proved vulnerable to adversarial texts. In this paper, we present RoChBERT, a framework to build more Robust BERT-based models by utilizing a more comprehensive adversarial graph to fuse Chinese phonetic and glyph features into pre-trained representations during fine-tuning. Inspired by curriculum learning, we further propose to augment the training dataset with adversarial texts in combination with intermediate samples. Extensive experiments demonstrate that RoChBERT outperforms previous methods in significant ways: (i) robust -- RoChBERT greatly improves the model robustness without sacrificing accuracy on benign texts. Specifically, the defense lowers the success rates of unlimited and limited attacks by 59.43% and 39.33% respectively, while remaining accuracy of 93.30%; (ii) flexible -- RoChBERT can easily extend to various language models to solve different downstream tasks with excellent performance; and (iii) efficient -- RoChBERT can be directly applied to the fine-tuning stage without pre-training language model from scratch, and the proposed data augmentation method is also low-cost.

cs.CL↗

arXiv2022

Annotating Columns with Pre-trained Language Models

Inferring meta information about tables, such as column headers or relationships between columns, is an active research topic in data management as we find many tables are missing some of this information. In this paper, we study the problem of annotating table columns (i.e., predicting column types and the relationships between columns) using only information from the table itself. We develop a multi-task learning framework (called Doduo) based on pre-trained language models, which takes the entire table as input and predicts column types/relations using a single model. Experimental results show that Doduo establishes new state-of-the-art performance on two benchmarks for the column type prediction and column relation prediction tasks with up to 4.0% and 11.9% improvements, respectively. We report that Doduo can already outperform the previous state-of-the-art performance with a minimal number of tokens, only 8 tokens per column. We release a toolbox (https://github.com/megagonlabs/doduo) and confirm the effectiveness of Doduo on a real-world data science problem through a case study.

cs.DB↗