With the rapid advancement of large language models (LLMs), producing realistic fake news has become increasingly effortless, challenging existing detection methods that rely on lexical and syntactic patterns. To address this, we shift our focus to the generation process and analyze how malicious prompts manipulate model outputs. We construct pairs of LLM-generated real and fake news and apply malicious prompts to reconstruct them as fake. By comparing the original-token generation probabilities recorded during reconstruction, we observe a consistent statistical divergence: tokens from real news tend to have lower reconstruction likelihoods than those from fake news. We define this distributional divergence as linguistic fingerprint. Building on this insight, we propose LIFE (Linguistic Fingerprints Extraction), a novel detection framework that reconstructs token-level probability distributions guided by malicious prompts to capture these discriminative linguistic patterns. To fully exploit the extracted fingerprints, LIFE further introduces a key-fragment amplification module that adaptively identifies and accentuates the most distinctive linguistic fragments, thereby enhancing detection reliability across diverse prompting scenarios. Extensive experiments demonstrate that LIFE achieves state-of-the-art performance in detecting LLM-generated fake news while maintaining strong generalization to human-LLM mixed cases. The code is available.

The rapid proliferation of fake news on social media threatens social stability, creating an urgent demand for more effective detection methods. While many promising approaches have emerged, most rely on content analysis with limited semantic depth, leading to suboptimal comprehension of news content. To address this limitation, capturing broader-range semantics is essential yet challenging, as it introduces two primary types of noise: fully connecting sentences in news graphs often adds unnecessary structural noise, while highly similar but authenticity-irrelevant sentences introduce feature noise, complicating the detection process. To tackle these issues, we propose BREAK, a broad-range semantics model for fake news detection that leverages a fully connected graph to capture comprehensive semantics while employing dual denoising modules to minimize both structural and feature noise. The semantic structure denoising module balances the graph's connectivity by iteratively refining it between two bounds: a sequence-based structure as a lower bound and a fully connected graph as the upper bound. This refinement uncovers label-relevant semantic interrelations structures. Meanwhile, the semantic feature denoising module reduces noise from similar semantics by diversifying representations, aligning distinct outputs from the denoised graph and sequence encoders using KL-divergence to achieve feature diversification in high-dimensional space. The two modules are jointly optimized in a bi-level framework, enhancing the integration of denoised semantics into a comprehensive representation for detection. Extensive experiments across four datasets prove that BREAK significantly outperforms existing fake news detection methods.

The wide dissemination of fake news has affected our lives in many aspects, making fake news detection important and attracting increasing attention. Existing approaches make substantial contributions in this field by modeling news from a single-modal or multi-modal perspective. However, these modal-based methods can result in sub-optimal outcomes as they ignore reader behaviors in news consumption and authenticity verification. For instance, they haven't taken into consideration the component-by-component reading process: from the headline, images, comments, to the body, which is essential for modeling news with more granularity. To this end, we propose an approach ofEmulating thebehaviors ofreaders (Ember) for fake news detection on social media, incorporating readers' reading and verificating process to model news from the component perspective thoroughly. Specifically, we first construct intra-component feature extractors to emulate the behaviors of semantic analyzing on each component. Then, we design a module that comprises inter-component feature extractors and a sequence-based aggregator. This module mimics the process of verifying the correlation between components and the overall reading and verification sequence. Thus, Ember can handle the news with various components by emulating corresponding sequences. We conduct extensive experiments on nine real-world datasets, and the results demonstrate the superiority of Ember.

In recent years, fake news on social media has become a significant threat to societal security, elevating fake news detection to a research priority. Among various strategies, fact-checking detection methods stand out for their accuracy, leveraging evidence from dedicated fact databases. However, these methods often retrieve raw truth, including vast amounts of irrelevant data, based on semantic similarity. This approach results in information redundancy and risks missing the nuanced differences between fake news and the truth. As a result, subtle changes in fake news can greatly increase the risk of misclassification, compromising the methods’ robustness. To this end, we propose a robust fake news detection framework with Fine-grained Discrepancy Contrastive Learning (FinDCL). By simulating subtle discrepancies between fake news and event-related truth, our method enhances the capture and identification of nuanced falsehoods. Specifically, we construct an adversarial dataset to pre-train a fine-grained discrepancy calculation module with contrastive learning. Moreover, the truth extraction module is devised to alleviate information redundancy by extracting event-related truth. At last, FinDCL jointly utilizes the aforementioned modules to detect fake news in event truth-known and truth-unknown scenarios. Extensive experiments on six real-world datasets demonstrate the effectiveness of FinDCL.

This work presents the design and implementation of TyrLoc, an accurate multi-technology switching MIMO localization system that can be deployed on low-cost SDRs. TyrLoc only uses a single RF Chain to switch on each antenna in an antenna array within the coherence time asynchronously, thus mimicking a MIMO platform to pinpoint the positions of WIFI, Bluetooth Low Energy (BLE) and LoRa devices. TyrLoc makes three key technical contributions. First, TyrLoc modifies the firmware of inexpensive PlutoSDR that controls the antenna switching pattern and tags the signal associated with each antenna. Second, it develops a two-stage fine-grained carrier frequency offset (CFO) calibration algorithm that harnesses the agile antenna switching pattern and is 10× more accurate than the baseline method. Third, TyrLoc employs an interpolated transform approach to facilitate angle-of-arrival (AoA) estimation in the presence of missing antennas. The AoA-based localization experiments in a multipath-rich indoor environment show that TyrLoc with eight antennas achieves the median errors of 63cm for WIFI, 39cm for BLE and 32cm for LoRa, respectively.