arXiv subjects

Yingtao Ren

Publications and source records attributed to Yingtao Ren.

4 recordsLinked to original sources

arXiv2026

When Context Bites: Detecting RAG Poisoning via Document-Level Attention Collapse

Retrieval-augmented generation (RAG) is indispensable for enhancing large language models. However, RAGs are increasingly susceptible to poisoning attacks, in which adversarial documents are injected to manipulate generator outputs. Previous methods rely on output-side signals such as perplexity and consistency checks to detect such attacks. Nevertheless, our analysis reveals that deliberate attacks often induce false confidence, where poisoned outputs exhibit even lower perplexity than benign ones, rendering uncertainty-based detection ineffective. To address this challenge, we explore the internal dynamics of the generator and identify a distinctive signature termed \textit{Attention Collapse}. Unlike the dispersed attention in benign generations, attacked generations exhibit a decrease in entropy as attention concentrates on poisoned documents. Building on these findings, we propose \texttt{D-SCAN} (Document-level Signal Collapse Analysis), a lightweight detection framework that monitors attention dynamics to identify attacked generations. Extensive experiments on multiple attack benchmarks demonstrate the effectiveness of our method. Moreover, D-SCAN can detect attacks even when they fail to alter the final answer. Code is available at https://github.com/yingtaoren/D-Scan.git.

cs.CR↗

arXiv2026

iFuzz-Meta: An Interpretable Fuzzy Learning Framework Bridging Top-Down and Bottom-Up Knowledge Integration

Interpretable representation learning remains a key challenge in modern neural computation, particularly when models are expected not only to perform but also to explain their reasoning. This paper introduces iFuzz-Meta, an interpretable fuzzy rule-based learning framework that preserves human-understandable reasoning structures within modern neural architectures. Each fuzzy rule corresponds to a semantic and spatial prototype defined in the original feature space, enabling transparent inference and direct interpretability. Meta-learning is employed as an analytical paradigm to examine how these interpretable rules reorganize across tasks and domains, providing a principled means to link algorithmic adaptation with cognitive representation. A knowledge-guided regularization mechanism further enables a top-down-bottom-up integration, in which theoretical priors act as soft inductive biases while data-driven learning refines and extends them. This dual process ensures that adaptation proceeds along semantically and physiologically meaningful trajectories, rather than arbitrary parameter shifts. Evaluations demonstrate that iFuzz-Meta achieves interpretable reasoning and stable cross-domain generalization, establishing a potential general pathway toward explainable and knowledge-aware fuzzy systems.

cs.LG↗

arXiv2024

A Self-Constructing Multi-Expert Fuzzy System for High-dimensional Data Classification

Fuzzy Neural Networks (FNNs) are effective machine learning models for classification tasks, commonly based on the Takagi-Sugeno-Kang (TSK) fuzzy system. However, when faced with high-dimensional data, especially with noise, FNNs encounter challenges such as vanishing gradients, excessive fuzzy rules, and limited access to prior knowledge. To address these challenges, we propose a novel fuzzy system, the Self-Constructing Multi-Expert Fuzzy System (SOME-FS). It combines two learning strategies: mixed structure learning and multi-expert advanced learning. The former enables each base classifier to effectively determine its structure without requiring prior knowledge, while the latter tackles the issue of vanishing gradients by enabling each rule to focus on its local region, thereby enhancing the robustness of the fuzzy classifiers. The overall ensemble architecture enhances the stability and prediction performance of the fuzzy system. Our experimental results demonstrate that the proposed SOME-FS is effective in high-dimensional tabular data, especially in dealing with uncertainty. Moreover, our stable rule mining process can identify concise and core rules learned by the SOME-FS.

cs.LG↗

arXiv2024

A New Random Forest Ensemble of Intuitionistic Fuzzy Decision Trees

Classification is essential to the applications in the field of data mining, artificial intelligence, and fault detection. There exists a strong need in developing accurate, suitable, and efficient classification methods and algorithms with broad applicability. Random forest is a general algorithm that is often used for classification under complex conditions. Although it has been widely adopted, its combination with diverse fuzzy theory is still worth exploring. In this paper, we propose the intuitionistic fuzzy random forest (IFRF), a new random forest ensemble of intuitionistic fuzzy decision trees (IFDT). Such trees in forest use intuitionistic fuzzy information gain to select features and consider hesitation in information transmission. The proposed method enjoys the power of the randomness from bootstrapped sampling and feature selection, the flexibility of fuzzy logic and fuzzy sets, and the robustness of multiple classifier systems. Extensive experiments demonstrate that the IFRF has competitative and superior performance compared to other state-of-the-art fuzzy and ensemble algorithms. IFDT is more suitable for ensemble learning with outstanding classification accuracy. This study is the first to propose a random forest ensemble based on the intuitionistic fuzzy theory.

cs.AI↗