Text generated by Large Language Models (LLMs) now rivals human writing, raising concerns about its misuse. However, mainstream AI-generated text detection (AGTD) methods primarily target document-level long texts and struggle to generalize effectively to sentence-level short texts. And current sentence-level AGTD (S-AGTD) research faces two significant limitations: (1) lack of a comprehensive evaluation on complex human-AI hybrid content, where human-written text (HWT) and AI-generated text (AGT) alternate irregularly, and (2) failure to incorporate contextual information, which serves as a crucial supplementary feature for identifying the origin of the detected sentence. Therefore, in our work, we propose AutoFill-Refine, a high-quality synthesis strategy for human-AI hybrid texts, and then construct a dedicated S-AGTD benchmark dataset. Besides, we introduce SenDetEX, a novel framework for sentence-level AI-generated text detection via style and context fusion. Extensive experiments demonstrate that SenDetEX significantly outperforms all baseline models in detection accuracy, while exhibiting remarkable transferability and robustness. Source code is available at https://github.com/TristoneJiang/SenDetEX.

Ensuring that Large Language Models (LLMs) return just responses which adhere to societal values is crucial for their broader application. Prior research has shown that LLMs often fail to perform satisfactorily on tasks requiring moral cognizance, such as ethics-based judgments. While current approaches have focused on fine-tuning LLMs with curated datasets to improve their capabilities on such tasks, choosing the optimal learning paradigm to enhance the ethical responses of LLMs remains an open research debate. In this work, we aim to address this fundamental question: can current learning paradigms enable LLMs to acquire sufficient moral reasoning capabilities? Drawing from distributional semantics theory and the pragmatic nature of moral discourse, our analysis indicates that performance improvements follow a mechanism similar to that of semantic-level tasks, and therefore remain affected by the pragmatic nature of morals latent in discourse, a phenomenon we name the pragmatic dilemma. We conclude that this pragmatic dilemma imposes significant limitations on the generalization ability of current learning paradigms, making it the primary bottleneck for moral reasoning acquisition in LLMs.

Packing for Supervised Fine-Tuning (SFT) in autoregressive models involves concatenating data points of varying lengths until reaching the designed maximum length to facilitate GPU processing. However, randomly concatenating data points can lead to cross-contamination of sequences due to the significant difference in their subject matter. The mainstream approaches in SFT ensure that each token in the attention calculation phase only focuses on tokens within its own short sequence, without providing additional learning signals for the preceding context. To address these challenges, we introduce Threshold Filtering Packing (TFP), a method that selects samples with related context while maintaining sufficient diversity within the same pack. Our experiments show that TFP offers a simple-to-implement and scalable approach that significantly enhances SFT performance, with observed improvements of up to 7% on GSM8K, 4% on HumanEval. Furthermore, results from bias benchmark datasets highlight TFP’s promising performance in improving fairness while also boosting prediction accuracy by 15%.

Prompt tuning is one of the most effective solutions to adapting a fixed pre-trained language model (PLM) for various downstream tasks, especially with only a few input samples. However, the security issues, e.g., Trojan attacks, of prompt tuning on a few data samples are not well-studied. Transferring established data poisoning attacks directly to few-shot prompt tuning presents multiple challenges. One significant issue is the _poisoned imbalance issue_, where non-target class samples are added to the target class, resulting in a greater number of target-class samples compared to non-target class. While this issue is not critical in regular tuning, it significantly hampers the few-shot prompt tuning, making it difficult to simultaneously achieve a high attack success rate (ASR) and maintain clean data accuracy (CDA). Additionally, few-shot prompting is prone to overfitting in terms of both ASR and CDA. In this paper, we introduce _TrojFSP_, a method designed to address the challenges. To solve the poisoned imbalance issue, we develop a _Target-Class Shrink (TC-Shrink)_ technique, which aims to equalize the number of poisoning samples. To combat overfitting, we employ a _Selective Token Poisoning_ technique to boost attack performance. Furthermore, we introduce a _Trojan-Trigger Attention_ objective function to amplify the attention of the poisoned trojan prompt on triggers. Experiments show that our TrojFSP achieves an ASR of over 99% while maintaining negligible decreases in CDA across various PLMs and datasets. The source code of TrojFSP is available at _https://github.com/UCF-ML-Research/TrojFSP_.

Data augmentation with mixup has shown to be effective on the NLP tasks. Although its great success, the mixup still has shortcomings. First, vanilla mixup randomly selects one sample to generate the mixup sample for a given sample. It remains unclear how to best choose the input samples for the mixup. Second, linear interpolation limits the space of synthetic data and its regularization effect. In this paper, we propose the dynamic nonlinear mixup with distance-based sample selection, which not only generates multiple sample pairs based on the distance between each sample but also enlarges the space of synthetic samples. Specifically, we compute the distance between each input data by cosine similarity and select multiple samples for a given sample. Then we use the dynamic nonlinear mixup to fuse sample pairs. It does not use a linear, scalar mixing strategy, but a nonlinear interpolation strategy, where the mixing strategy is adaptively updated for the input and label pairs. Experiments on the multiple public datasets demonstrate that dynamic nonlinear mixup outperforms state-of-the-art methods.

Aspect-based sentiment analysis (ABSA) is a fine-grained sentiment classification task. Most recent efforts adopt pre-trained model to classify the sentences with aspects. However, the aspect sentiment bias from pre-trained model brings some noise to the ABSA task. Besides, traditional methods using cross-entropy loss are hard to find the potential associations between sentiment polarities. In this work, we analyze the ABSA task from a novel cognition perspective: humans can often judge the sentiment of an aspect even if they do not know what the aspect is. Moreover, it is easier to distinguish positive and negative sentiments than others for human beings because positive and negative are two opposite sentiments. To this end, we propose a no-aspect differential sentiment (NADS) framework for the ABSA task. We first design a no-aspect template by replacing the aspect with a special unbiased character to eliminate the sentiment bias and obtain a stronger representation. To better get the benefits from the template, we adopt contrastive learning between the no-aspect template and the original sentence. Then we propose a differential sentiment loss instead of the cross-entropy loss to better classify the sentiments by distinguishing the different distances between sentiments. Our proposed model is a general framework and can be combined with almost all traditional ABSA methods. Experiments on SemEval 2014 show that our framework is still able to predict the sentiment of the aspect even we don’t konw what the aspect is. Moreover, our NADS framework boosts three typical ABSA methods and achieves state-of-the-art performance.

Homomorphic encryption (HE) and garbled circuit (GC) provide the protection for users’ privacy. However, simply mixing the HE and GC in RNN models suffer from long inference latency due to slow activation functions. In this paper, we present a novel hybrid structure of HE and GC gated recurrent unit (GRU) network, , for low-latency secure inferences. replaces computationally expensive GC-based tanh with fast GC-based ReLU, and then quantizes sigmoid and ReLU to smaller bit-length to accelerate activations in a GRU. We evaluate with multiple GRU models trained on 4 public datasets. Experimental results show achieves top-notch accuracy and improves the secure inference latency by up to 138× over one of the state-of-the-art secure networks on the Penn Treebank dataset.

We propose a method called reverse mapping bytepair encoding, which maps named-entity information and other word-level linguistic features back to subwords during the encoding procedure of bytepair encoding (BPE). We employ this method to the Generative Pre-trained Transformer (OpenAI GPT) by adding a weighted linear layer after the embedding layer. We also propose a new model architecture named as the multi-channel separate transformer to employ a training process without parameter-sharing. Evaluation on Stories Cloze, RTE, SciTail and SST-2 datasets demonstrates the effectiveness of our approach.