Ziyi Zhu

Robert Irvine, Douglas Boubert, Vyas Raina et al. · cambridge

The emergence of pretrained large language models has led to the deployment of a range of social chatbots for chitchat. Although these chatbots demonstrate language ability and fluency, they are not guaranteed to be engaging and can struggle to retain users. This work investigates the development of social chatbots that prioritize user engagement to enhance retention, specifically examining the use of human feedback to efficiently develop highly engaging chatbots. The proposed approach uses automatic pseudo-labels collected from user interactions to train a reward model that can be used to reject low-scoring sample responses generated by the chatbot model at inference time. Intuitive evaluation metrics, such as mean conversation length (MCL), are introduced as proxies to measure the level of engagement of deployed chatbots. A/B testing on groups of 10,000 new daily chatbot users on the Chai Research platform shows that this approach increases the MCL by up to 70%, which translates to a more than 30% increase in user retention for a GPT-J 6B model. Future work aims to use the reward model to realise a data fly-wheel, where the latest user conversations can be used to alternately fine-tune the language model and the reward model.

Ziyi Zhu, Olivier Tieleman, Caitlin A. Stamatis et al. · cambridge

Realistic user simulation is crucial for training and evaluating task-oriented dialogue (TOD) systems, yet creating simulators that accurately replicate human behavior remains challenging. A key property of effective simulators is their ability to expose failure modes of the systems they evaluate. We present an adversarial training framework that iteratively improves user simulator realism through a competitive dynamic between a generator (user simulator) and a discriminator. Applied to mental health support chatbots, our approach demonstrates that fine-tuned simulators dramatically outperform zero-shot base models at surfacing system issues, and adversarial training further enhances diversity, distributional alignment, and predictive validity. The resulting simulator achieves a strong correlation between simulated and real failure occurrence rates across diverse chatbot configurations while maintaining low distributional divergence of failure modes. Discriminator accuracy decreases drastically after three adversarial iterations, suggesting improved realism. These results provide evidence that adversarial training is a promising approach for creating realistic user simulators in mental health support TOD domains, enabling rapid, reliable, and cost-effective system evaluation before deployment.

Ziyi Zhu, Olivier Tieleman, Alexey Bukhtiyarov et al. · cambridge

LLM-as-judge evaluation has become standard practice for open-ended model assessment; however, judges exhibit systematic biases that cannot be eliminated by increasing the number of scenarios or generations. These biases are often similar in magnitude to the model differences that benchmarks are designed to detect, resulting in unreliable rankings when single-judge evaluations are used. This work introduces a variance decomposition that partitions benchmark score variance into scenario, generation, judge, and residual components. Based on this analysis, CyclicJudge, a round-robin assignment of judges, is demonstrated to be the optimal allocation strategy. It eliminates bias precisely while requiring each judge only once per cycle, maintaining the cost of single-judge evaluation. Empirical validation on MT-Bench supports all theoretical predictions.

Yuhong Mo, Hao Qin, Yushan Dong et al.

In this paper, a tool for detecting LLM AI text generation is developed based on the Transformer model, aiming to improve the accuracy of AI text generation detection and provide reference for subsequent research. Firstly the text is Unicode normalised, converted to lowercase form, characters other than non-alphabetic characters and punctuation marks are removed by regular expressions, spaces are added around punctuation marks, first and last spaces are removed, consecutive ellipses are replaced with single spaces and the text is connected using the specified delimiter. Next remove non-alphabetic characters and extra whitespace characters, replace multiple consecutive whitespace characters with a single space and again convert to lowercase form. The deep learning model combines layers such as LSTM, Transformer and CNN for text classification or sequence labelling tasks. The training and validation sets show that the model loss decreases from 0.127 to 0.005 and accuracy increases from 94.96 to 99.8, indicating that the model has good detection and classification ability for AI generated text. The test set confusion matrix and accuracy show that the model has 99% prediction accuracy for AI-generated text, with a precision of 0.99, a recall of 1, and an f1 score of 0.99, achieving a very high classification accuracy. Looking forward, it has the prospect of wide application in the field of AI text detection.

Lingxi Xiao, Muqing Li, Yinqiu Feng et al.

The research explores the utilization of a deep learning model employing an attention mechanism in medical text mining. It targets the challenge of analyzing unstructured text information within medical data. This research seeks to enhance the model's capability to identify essential medical information by incorporating deep learning and attention mechanisms. This paper reviews the basic principles and typical model architecture of attention mechanisms and shows the effectiveness of their application in the tasks of disease prediction, drug side effect monitoring, and entity relationship extraction. Aiming at the particularity of medical texts, an adaptive attention model integrating domain knowledge is proposed, and its ability to understand medical terms and process complex contexts is optimized. The experiment verifies the model's effectiveness in improving task accuracy and robustness, especially when dealing with long text. The future research path of enhancing model interpretation, realizing cross-domain knowledge transfer, and adapting to low-resource scenarios is discussed in the research outlook, which provides a new perspective and method support for intelligent medical information processing and clinical decision assistance. Finally, cross-domain knowledge transfer and adaptation strategies for low-resource scenarios, providing theoretical basis and technical reference for promoting the development of intelligent medical information processing and clinical decision support systems.

Yang Lin, Muqing Li, Ziyi Zhu et al.

The prediction of disease risk factors can screen vulnerable groups for effective prevention and treatment, so as to reduce their morbidity and mortality. Machine learning has a great demand for high-quality labeling information, and labeling noise in medical big data poses a great challenge to efficient disease risk warning methods. Therefore, this project intends to study the robust learning algorithm and apply it to the early warning of infectious disease risk. A dynamic truncated loss model is proposed, which combines the traditional mutual entropy implicit weight feature with the mean variation feature. It is robust to label noise. A lower bound on training loss is constructed, and a method based on sampling rate is proposed to reduce the gradient of suspected samples to reduce the influence of noise on training results. The effectiveness of this method under different types of noise was verified by using a stroke screening data set as an example. This method enables robust learning of data containing label noise.