The proliferation of Large Language Models (LLMs) such as ChatGPT and Gemini has resulted in a surge of AI-generated text across various domains. However, the widespread use of this technology raises concerns regarding the generation of misinformation and malicious content. To address this challenge, we propose a novel AI-generated Text Detection model combining Probabilistic and Semantic features (ATDPS). Our model extracts semantic features using a pre-trained language model and combines them with probabilistic features generated by multiple LLMs. A temporal convolutional network is employed to process sequence probabilistic features, effectively capturing temporal characteristics within the text. To ensure data coherence and diversity, our dataset includes text generated by a variety of LLMs, including the latest models like GPT-4. Experimental results demonstrate ATDPS's superior performance over existing baselines in terms of accuracy, precision, recall and F1 score, highlighting its potential and effectiveness in detecting AI-generated text.
Yang Yu, Wang Gao· International Journal of Sci...· 0 citations
Generative image editing models struggle with structured statistical charts when data modifications require geometric synchronization. We formalize this task as Visuo-Logical Cascading Editing (VLCE). However, existing methods remain confined to localized text substitutions and struggle with dependency-aware cascading updates. To systematically evaluate this capability, we introduce ChartSync, an expert-validated benchmark constructed via a programmatic rendering pipeline that guarantees deterministic visuo-logical coupling for the ground truth. ChartSync comprises 870 triplets across 9 chart categories and 4 task types, including 235 geometry-coupled VLCE instances that specifically test cascading text-to-geometry synchronization. We further evaluate these instances via a two-tier framework combining objective visual metrics with a vision-language model judge paradigm to assess low-level fidelity alongside multimodal comprehension and reasoning. Evaluating 14 image editing models and one code-mediated pipeline reveals a nuanced capability gap: most open-source models suffer severe drops in geometric synchronization, while only two frontier proprietary models show emerging VLCE capability, with their residual errors mainly involving semantic isolation and background corruption. Our detailed error analysis deconstructs these failure paradigms to identify core meta-abilities for guiding future multimodal architectures. The ChartSync dataset and code are publicly released at https://github.com/kaka-yjk/ChartSyncCodebase.
Jiakang Yu, Yixuan Chai, Tianci Wang et al.· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
