arXiv subjects

Liting Li

Publications and source records attributed to Liting Li.

4 recordsLinked to original sources

arXiv2026

TRACE: Trajectory Risk-Aware Compression for Long-Horizon Agent Safety

Long-horizon LLM agents produce safety evidence across long trajectories, where sparse, delayed, and compositional risk signals often escape local moderation. Existing turn-level or short-context detectors struggle to reliably retain and aggregate such evidence over extended horizons. We reframe long-horizon agent safety detection as trajectory-level evidence compression and propose Trajectory Risk-Aware Compression for Long-Horizon Agent Safety (TRACE). TRACE uses a Compressor-Reader design: the Compressor encodes the full trajectory into a compact latent evidence state under trajectory-level supervision, and the Reader judges the raw trajectory with this latent evidence state as a safety reference. This design helps aggregate dispersed risk cues and reduce premature evidence loss. Across ASSEBench, Pre-Ex-Bench, and R-Judge, TRACE achieves the best accuracy on all evaluated backbones, improving over strong baselines by up to 12.6 percentage points. On LongSafety, TRACE shows smaller performance degradation as context length grows. Attention visualizations and case studies suggest that the compressed reference helps the Reader focus on risk-critical segments and recover cross-step evidence. Code is available at https://github.com/Peregrine123/TRACE_official.

cs.AI↗

arXiv2025

Meta-Learning Based Few-Shot Graph-Level Anomaly Detection

Graph-level anomaly detection aims to identify anomalous graphs or subgraphs within graph datasets, playing a vital role in various fields such as fraud detection, review classification, and biochemistry. While Graph Neural Networks (GNNs) have made significant progress in this domain, existing methods rely heavily on large amounts of labeled data, which is often unavailable in real-world scenarios. Additionally, few-shot anomaly detection methods based on GNNs are prone to noise interference, resulting in poor embedding quality and reduced model robustness. To address these challenges, we propose a novel meta-learning-based graph-level anomaly detection framework (MA-GAD), incorporating a graph compression module that reduces the graph size, mitigating noise interference while retaining essential node information. We also leverage meta-learning to extract meta-anomaly information from similar networks, enabling the learning of an initialization model that can rapidly adapt to new tasks with limited samples. This improves the anomaly detection performance on target graphs, and a bias network is used to enhance the distinction between anomalous and normal nodes. Our experimental results, based on four real-world biochemical datasets, demonstrate that MA-GAD outperforms existing state-of-the-art methods in graph-level anomaly detection under few-shot conditions. Experiments on both graph anomaly and subgraph anomaly detection tasks validate the framework's effectiveness on real-world datasets.

cs.LG↗

arXiv2025

Empirical Validation of Functional Multidimensional Scaling via Numerical Simulation and Real-World Application

This article presents an empirical validation of the functional multidimensional scaling model, a novel approach that improves the smoothness of time-varying dissimilarities in a low-dimensional space, embedding a modified Adam stochastic gradient descent method. We conduct a numerical simulation study to evaluate the feasibility of the functional multidimensional scaling model under various controlled scenarios and to assess the goodness of the approximation of the estimators with a curvilinear search method, demonstrating its robustness and scalability in dynamic structures. To further explore its effectiveness in practice, we implement the functional multidimensional scaling model in a real-world case with stock market data, revealing strong clustering capabilities in visualization. The experiments in this article indicate that the functional multidimensional scaling model performs robustly on synthetic benchmarks and provides meaningful insights of high-dimensional and time-varying data in the real world, reinforcing its value in practical applications.

stat.AP↗

arXiv2025

Functional Multidimensional Scaling

This article introduces a functional method for lower-dimensional smooth representations in terms of time-varying dissimilarities. The method incorporates dissimilarity representation in multidimensional scaling and smoothness approach of functional data analysis by using cubic B-spline basis functions. The model is designed to arrive at optimal representations with an iterative procedure such that dissimilarities evaluated by estimated representations are almost the same as original dissimilarities of objects in a low dimension which is easier for people to recognize. To solve expensive computation in optimization, we propose a computationally efficient method by taking gradient steps with respect to individual sub-functions of target functions using a Stochastic Gradient Descent algorithm. Keywords: Multidimensional Scaling, Functional Data Analysis, Statistical Modeling, Quasi-Newton Method, Stochastic Gradient Descent

stat.ME↗