Recent advances in generative AI technologies like large language models have boosted the incorporation of AI assistance in writing workflows, leading to the rise of a new paradigm of human-AI co-creation in writing. To understand how people perceive writings that are produced under this paradigm, in this paper, we conduct an experimental study to understand whether and how the disclosure of the level and type of AI assistance in the writing process would affect people’s perceptions of the writing on various aspects, including their evaluation on the quality of the writing, and their ranking of different writings. Our results suggest that disclosing the AI assistance in the writing process, especially if AI has provided assistance in generating new content, decreases the average quality ratings for both argumentative essays and creative stories. This decrease in the average quality ratings often comes with an increased level of variations in different individuals’ quality evaluations of the same writing. Indeed, factors such as an individual’s writing confidence and familiarity with AI writing assistants are shown to moderate the impact of AI assistance disclosure on their writing quality evaluations. We also find that disclosing the use of AI assistance may significantly reduce the proportion of writings produced with AI’s content generation assistance among the top-ranked writings.
Experts across diverse disciplines are often interested in making sense of large text collections. Traditionally, this challenge is approached either by noisy unsupervised techniques such as topic models, or by following a manual theme discovery process. In this paper, we expand the definition of a theme to account for more than just a word distribution, and include generalized concepts deemed relevant by domain experts. Then, we propose an interactive framework that receives and encodes expert feedback at different levels of abstraction. Our framework strikes a balance between automation and manual coding, allowing experts to maintain control of their study while reducing the manual effort required.
The recent LLMs like GPT-4 and PaLM-2 have made tremendous progress in solving fundamental math problems like GSM8K by achieving over 90% accuracy. However, their capabilities to solve more challenging math problems which require domain-specific knowledge (i.e. theorem) have yet to be investigated. In this paper, we introduce TheoremQA, the first theorem-driven question-answering dataset designed to evaluate AI models’ capabilities to apply theorems to solve challenging science problems. TheoremQA is curated by domain experts containing 800 high-quality questions covering 350 theorems from Math, Physics, EE&CS, and Finance. We evaluate a wide spectrum of 16 large language and code models with different prompting strategies like Chain-of-Thoughts and Program-of-Thoughts. We found that GPT-4’s capabilities to solve these problems are unparalleled, achieving an accuracy of 51% with Program-of-Thoughts Prompting. All the existing open-sourced models are below 15%, barely surpassing the random-guess baseline. Given the diversity and broad coverage of TheoremQA, we believe it can be used as a better benchmark to evaluate LLMs’ capabilities to solve challenging science problems.
The collection and curation of high-quality training data is crucial for developing text classification models with superior performance, but it is often associated with significant costs and time investment. Researchers have recently explored using large language models (LLMs) to generate synthetic datasets as an alternative approach. However, the effectiveness of the LLM-generated synthetic data in supporting model training is inconsistent across different classification tasks. To better understand factors that moderate the effectiveness of the LLM-generated synthetic data, in this study, we look into how the performance of models trained on these synthetic data may vary with the subjectivity of classification. Our results indicate that subjectivity, at both the task level and instance level, is negatively associated with the performance of the model trained on synthetic data. We conclude by discussing the implications of our work on the potential and limitations of leveraging LLM for synthetic data generation.
The Covid-19 pandemic has led to infodemic of low quality information leading to poor health decisions. Combating the outcomes of this infodemic is not only a question of identifying false claims, but also reasoning about the decisions individuals make. In this work we propose a holistic analysis framework connecting stance and reason analysis, and fine-grained entity level moral sentiment analysis. We study how to model the dependencies between the different level of analysis and incorporate human insights into the learning process. Experiments show that our framework provides reliable predictions even in the low-supervision settings.
Automated methods for analyzing public opinion have grown in popularity with the proliferation of social media. While supervised methods can be very good at classifying text, the dynamic nature of social media discourse results in a moving target for supervised learning. Meanwhile, traditional unsupervised techniques for extracting themes from textual repositories, such as topic models, can result in incorrect outputs that are unusable to domain experts. For this reason, a non-trivial amount of research on social media discourse still relies on manual coding techniques. In this paper, we present an interactive, humans-in-the-loop framework that strikes a balance between unsupervised techniques and manual coding for extracting latent arguments from social media discussions. We use the COVID-19 vaccination debate as a case study, and show that our methodology can be used to obtain a more accurate, interpretable set of arguments when compared to traditional topic models. We do this at a relatively low manual cost, as 3 experts take approximately 2 hours to code close to 100k tweets.