We propose a voting-driven semi-supervised approach to automatically acquire the typical duration of an event and use it as pseudo-labeled data. The human evaluation demonstrates that our pseudo labels exhibit surprisingly high accuracy and balanced coverage. In the temporal commonsense QA task, experimental results show that using only pseudo examples of 400 events, we achieve performance comparable to the existing BERT-based weakly supervised approaches that require a significant amount of training examples. When compared to the RoBERTa baselines, our best approach establishes state-of-the-art performance with a 7% improvement in Exact Match.
The creation of instruction data and evaluation benchmarks for serving Large language models often involves enormous human annotation. This issue becomes particularly pronounced when rapidly developing such resources for a non-English language like Japanese. Instead of following the popular practice of directly translating existing English resources into Japanese (e.g., Japanese-Alpaca), we propose an efficient self-instruct method based on GPT-4. We first translate a small amount of English instructions into Japanese and post-edit them to obtain native-level quality. GPT-4 then utilizes them as demonstrations to automatically generate Japanese instruction data. We also construct an evaluation benchmark containing 80 questions across 8 categories, using GPT-4 to automatically assess the response quality of LLMs without human references. The empirical results suggest that the models fine-tuned on our GPT-4 self-instruct data significantly outperformed the Japanese-Alpaca across all three base pre-trained models. Our GPT-4 self-instruct data allowed the LLaMA 13B model to defeat GPT-3.5 (Davinci-003) with a 54.37\% win-rate. The human evaluation exhibits the consistency between GPT-4's assessments and human preference. Our high-quality instruction data and evaluation benchmark have been released here.
ChatGPT and other Generative Artificial Intelligence (GAI) models tend to inherit and even amplify prevailing societal biases as they are trained on large amounts of existing data. Given the increasing usage of ChatGPT and other GAI by students, faculty members, and staff in higher education institutions (HEIs), there is an urgent need to examine the ethical issues involved such as its potential biases. In this scoping review, we clarify the ways in which biases related to GAI in higher education settings have been discussed in recent academic publications and identify what type of potential biases are commonly reported in this body of literature. We searched for academic articles written in English, Chinese, and Japanese across four main databases concerned with GAI usage in higher education and bias. Our findings show that while there is an awareness of potential biases around large language models (LLMs) and GAI, the majority of articles touch on ``bias'' at a relatively superficial level. Few identify what types of bias may occur under what circumstances. Neither do they discuss the possible implications for the higher education, staff, faculty members, or students. There is a notable lack of empirical work at this point, and we call for higher education researchers and AI experts to conduct more research in this area.
This scoping review explores the ethical challenges of using ChatGPT in education, focusing particularly on issues related to higher education. By reviewing recent academic articles written in English, Chinese, and Japanese, we aimed to provide a comprehensive overview of relevant research while identifying gaps for future considerations. Drawing on Arksey and O'Malley's (2005) five-stage scoping review framework, we identified research questions, search terms, and conducted article search from four databases in the target three languages. Each article was reviewed by at least two researchers identifying the main ethical issues of utilizing AI in education, particularly higher education. Our analysis of ethical issues followed the framework developed by DeepMind (Weiginger et al., 2021) to identify six main areas of ethical concern in Language Models. The majority of papers were concerned with misinformation harms (n=25) and/or human-computer interaction related harms (n=24). Given the rapid deployment of Generative Artificial Intelligence (GAI), it is imperative for educators to conduct more empirical studies to develop sound ethical policies for the use of GAI.
Temporal relation classification is a pair-wise task for identifying the relation of a temporal link (TLINK) between two mentions, i.e. event, time, and document creation time (DCT). It leads to two crucial limits: 1) Two TLINKs involving a common mention do not share information. 2) Existing models with independent classifiers for each TLINK category (E2E, E2T, and E2D) hinder from using the whole data. This paper presents an event centric model that allows to manage dynamic event representations across multiple TLINKs. Our model deals with three TLINK categories with multi-task learning to leverage the full size of data. The experimental results show that our proposal outperforms state-of-the-art models and two transfer learning baselines on both the English and Japanese data.
While large language models (LLMs) like GPT-4 have recently demonstrated astonishing zero-shot capabilities in general domain tasks, they often generate content with hallucinations in specific domains such as Chinese law, hindering their application in these areas. This is typically due to the absence of training data that encompasses such a specific domain, preventing GPT-4 from acquiring in-domain knowledge. A pressing challenge is that it's not plausible to continue training LLMs of such scale on in-domain data. This paper introduces a simple and effective domain adaptation framework for GPT-4 by reformulating generation as an \textbf{adapt-retrieve-revise} process. The initial step is to \textbf{adapt} an affordable 7B LLM to the target domain by continuing learning on in-domain data. When solving a task, we leverage the adapted LLM to generate a draft answer given a task query. Then, the draft answer will be used to \textbf{retrieve} supporting evidence candidates from an external in-domain knowledge base. Finally, the draft answer and retrieved evidence are concatenated into a whole prompt to let GPT-4 assess the evidence and \textbf{revise} the draft answer to generate the final answer. Our proposal combines the advantages of the efficiency of adapting a smaller 7B model with the evidence-assessing capability of GPT-4 and effectively prevents GPT-4 from generating hallucinatory content. In the zero-shot setting of four Chinese legal tasks, our method improves accuracy by 33.3\% compared to the direct generation by GPT-4. When compared to two stronger retrieval-based baselines, our method outperforms them by 15.4\% and 23.9\%. Our code will be released
Despite the success of ChatGPT, its performances on most NLP tasks are still well below the supervised baselines. In this work, we looked into the causes, and discovered that its subpar performance was caused by the following factors: (1) token limit in the prompt does not allow for the full utilization of the supervised datasets; (2) mismatch between the generation nature of ChatGPT and NLP tasks; (3) intrinsic pitfalls of LLMs models, e.g., hallucination, overly focus on certain keywords, etc. In this work, we propose a collection of general modules to address these issues, in an attempt to push the limits of ChatGPT on NLP tasks. Our proposed modules include (1) a one-input-multiple-prompts strategy that employs multiple prompts for one input to accommodate more demonstrations; (2) using fine-tuned models for better demonstration retrieval; (3) transforming tasks to formats that are more tailored to the generation nature; (4) employing reasoning strategies that are tailored to addressing the task-specific complexity; (5) the self-verification strategy to address the hallucination issue of LLMs; (6) the paraphrase strategy to improve the robustness of model predictions. We conduct experiments on 21 datasets of 10 representative NLP tasks, including question answering, commonsense reasoning, natural language inference, sentiment analysis, named entity recognition, entity-relation extraction, event extraction, dependency parsing, semantic role labeling, and part-of-speech tagging. Using the proposed assemble of techniques, we are able to significantly boost the performance of ChatGPT on the selected NLP tasks, achieving performances comparable to or better than supervised baselines, or even existing SOTA performances.
Large language models (LLMs) have achieved impressive performance on various reasoning tasks. To further improve the performance, we propose MultiTool-CoT, a novel framework that leverages chain-of-thought (CoT) prompting to incorporate multiple external tools, such as a calculator and a knowledge retriever, during the reasoning process. We apply MultiTool-CoT to the Task 2 dataset of NumGLUE, which requires both numerical reasoning and domain-specific knowledge. The experiments show that our method significantly outperforms strong baselines and achieves state-of-the-art performance.
Numerical reasoning over table-and-text hybrid passages, such as financial reports, poses significant challenges and has numerous potential applications. Noise and irrelevant variables in the model input have been a hindrance to its performance. Additionally, coarse-grained supervision of the whole solution program has impeded the model's ability to learn the underlying numerical reasoning process. In this paper, we propose three pretraining tasks that operate at both the whole program and sub-program level: Variable Integrity Ranking, which guides the model to focus on useful variables; Variable Operator Prediction, which decomposes the supervision into fine-grained single operator prediction; and Variable Keyphrase Masking, which encourages the model to identify key evidence that sub-programs are derived from. Experimental results demonstrate the effectiveness of our proposed methods, surpassing transformer-based model baselines.